Annual report 2015 of Innovative Measurement and …...The First Symposium on SIP Innovative measurement and analysis for structural materials (SIP-IMASM 2015) 5 3-2 Kazuhiro Hono

Annual report 2015

of Innovative Measurement and Analysis for Structural Materials

SIP-IMASMInnovative measurement and analysis for structural materials

2015

1st Symposium on Innovative Measurement and

Analysis for Structural Materials Sept. 29 – Oct. 1, 2015 AIST Tsukuba Center

第 1 回革新的構造材料のための先端計測拠点国際会議

The SIP-IMASM is supported by the Structural Materials for Innovation (SM4I), Cross-ministerial Strategic Innovation Promotion Program (SIP). This booklet contains the abstracts of the keynote and the invited speakers in addition to the SIP-IMASM members.

AIST15-C00008

The First Symposium on SIP Innovative measurement and analysis for structural materials (SIP-IMASM 2015)

1

Day-1 Sept. 29 (Tue) Day-2 Sept. 3 (Wed) Day-3 Oct. 1 (Thu)

9:00 9:00 9:00

AIST

9:50 NIMS

Tsukuba Uni.10:20 10:20 KEK

10:30

10:50

11:00

11:50

12:00 12:00

13:20 13:20

14:20 14:20

15:20 15:20

15:40 15:40

16:10

16:20 16:40

18:0017:40

17:50

18:00

20:00

Lunch Lunch

Registration

Banquet

BreakM. Enoki

(SIP MI domain)

(30)

Poster

N. Ohshima

(IMASM-4)A. Uedono

(IMASM-4)K. Wada

(IMASM-4)closing

Y. Takeichi IMASM-D. Fujita

IMASM-3

Break Break

Y. Kagawa

(Keynote 2)

(30)

Y. Shirai

(Invite)

(30)

M. Kumura (Invited)

(30)

T. Kitashima

SIP Alloy Domain)

(30)

M. Kimura (IMASM-1)H. Kitazawa

(IMASM-3)

Y. Tanaka IMASM-K. Hono

(IMASM-3)

Break

S. Kitaoka

(SIP Ceramics Domain)

(30)

M. Ohnuma

(Invited)

(30)

M. IONESCU

(Invited)

(40+10)

Opening K. Fukutani

(Invited)

(30)M. Ohkubo

T. Kishi (Keynote 1)

(30)

Break

N. Takeda

(SIP Polymer Domain)

(30)Break

J. D. Almer (Invited)

(40+10)

E. Kita

(IMASM-2)

D. Sekiba

IMASM-2F. Paul

IMASM-2Photo

(Details are announced.)


2

9/29(Tue.)

09:00 Registration

Session chair: Paul Fons (AIST)

09:50 Guest speech Eizo Matsumoto (CAO)

Introduction Masataka Ohkubo (AIST) "Introduction to SIP-IMASM" 9

Session chair: Masataka Ohkubo (AIST)

10:20 Keynote Teruo Kishi (CAO) "Structural Materials for Innovation" 8

10:50 Coffee Break

11:00 Invited J. D. Almer (ANL) "XRD Analysis of Structure Material at APS" 14

11:50 Symposium Photo, Lunch

Session chair: Eiji Kita (U.Tsukuba)

13:20 Invited Satoshi Kitaoka (JFCC) "Development of Ceramic Environmental Barrier Coatings for

Advanced Airplane Engine Applications" 21

13:50 Invited Masao Kimura (KEK) "In situ Observation using Synchrotron Radiation of Various

Reactions in Steel Processes" 26

14:20 Theme-1 Masao Kimura (KEK) "Overview of theme 1: Imaging of Strain, Cracks and Chemical

States of Structural Materials" 23

14:40 Theme-1 Yoshihisa Tanaka (NIMS) "In-situ Multi-scale Strain Imaging for Composite Materials using

FE-SEM during Mechanical and Thermal Loading" 26

15:00 Theme-1 Yasuo Takeichi (KEK) “Application of XAFS & XRD Mapping Techniques for Various

Materials” 29

15:20 Coffee Break

15:40 Keynote Yutaka Kagawa (U. Tokyo) " Concept of "Materials Integration" and Some Examples" 33

16:20 Poster session (see page 4)

18:00 Banquet

1st Symposium on Innovative Measurement and Analysis for Structural Materials (SIP-IMASM2015)


3

Session chair: Hideaki Kitazawa(NIMS)

09:00 Invited M. Ionescu (ANSTO) "Ion Beam Analysis for Materials Science" 35

09:50 Invited Katsuyuki Fukutani (U.Tokyo) "Analysis of hydrogen in materials with the 15N nuclear

reaction combined with thermal desorption spectroscopy"

10:20 Coffee Break

10:30 Invited Nobuo Takeda (U.Tokyo) "Life Cycle Monitoring and Quality Assessment of Advanced

Polymer Matrix Composites" 39

11:00 Theme-2 Eiji Kita (U.Tsukuba) "Overview of theme 2: Quantitative Elemental Analyses of

Hydrogen and Light Elements in Structural Materials" 43

11:20 Theme-2 Daiichiro Sekiba (U.Tsukuba) "IBA for Hydrogen Uptake Observation in Functional Metals

under Ambient Condition" 45

11:40 Theme-2 Fons Paul (AIST) "XAFS Structural and Chemical Analysis Approach for the

Trace Light Elements of B and N in Heat-Resistant Steel" 46

12:00 Lunch

Session chair: Nagayasu Oshima (AIST)

13:20 Invited Masato Ohnuma (Hokkaido U.) "Nanoscale Characterization in Structural Materials by SAXS

and SANS" 50

13:50 Invited Tomonori Kitashima (NIMS) "Development of High Temperature Titanium Alloys,

Microstructure and Property Prediction Methods" 51

14:20 Theme-3 Hideaki Kitazawa(NIMS) " Overview of theme 3: Development of Multiscale

Characterization in Structural Materials" 57

14:40 Theme-3 Kazuhiro Hono (NIMS) " Microstructure Characterization of Structural Materials using

Laser Assisted 3D Atom Probe" 60

15:00 Theme-3 Daisuke Fujita (NIMS) "Nanoscale Characterization of Structural Composite Materials"

61

15:20 Coffee Break

Session chair: Masao Kimura (KEK)

15:40 Invited Yasuharu Shirai (Kyoto U.) "Positron Annihilation Study of Vacancy-Type Defects in Iron

and Steels" 63

16:10 Invited Manabu Enoki (U.Tokyo) "Development of Performance Prediction System" 64

16:40 Theme-4 Nagayasu Oshima (AIST) "Overview of theme 4: Positron Annihilation Spectroscopy Based

Research for the SIP-IMASM project" 66

17:00 Theme-4 Akira Uedono (U.Tsukuba) "Vacancy-Type Defects and Open Spaces in Solid-State

Materials Probed by Means of Positron Annihilation" 68

17:20 Theme-4 Ken Wada (KEK) "Performance Test of a Pulse Stretch System for Materials

Sciences at KEK Slow Positron Facility" 74

17:40 Closing


4

Theme-1

1-1 Masao Kimura (KEK) " Overview of theme 1: Imaging of Strain, Cracks and Chemical States of

Structural Materials " 23

1-2 Yoshihisa Tanaka (NIMS) "In-situ Multi-scale Strain Imaging for Composite Materials using FE-SEM

during Mechanical and Thermal Loading" 26

1-3 Yasuo Takeichi (KEK) "Application of XAFS & XRD Mapping Techniques for Various Materials"

1-4 Keiichi Hirano (KEK) " Development of X-ray Multiple Image Radiography at the Photon Factory " 78

1-5 Yasuhiro Niwa (KEK) "Development of Nanosecond Time-Resolved Dispersive XAFS System for

Irreversible Phenomena" 81

1-6 Yumiko Takahashi (KEK) "CT imaging of Structure Materials" 84

1-7 Qinghua Wang (AIST) "Full-Field Deformation Measurement of Carbon Fiber Reinforced Plastics under

Three-Point Bending Test at Micro Scale" 87

1-8 Makoto Watanabe (NIMS)"Laser ultrasonic testing of carbon-fiber-reinforced plastic (CFRP) with Mid IR

pulsed light source generated by OPO" 93

Theme-2

2-1 Eiji Kita (U.Tsukuba) "Overview of theme 2: Quantitative Elemental Analyses of Hydrogen and Light

Elements in Structural Materials" 43

2-2 Daiichiro Sekiba (U.Tsukuba) "IBA for Hydrogen Uptake Observation in Functional Metals under Ambient

Condition" 45

2-3 Fons Paul (AIST) "An XAFS Structural and Chemical Analysis Approach for the Trace Light

Elements of B and N in Heat-Resistant Steel" 46

2-4 Shigetomo Shiki (AIST) "Superconducting X-ray Spectrometer for Trace Light Elements in Structural

Materials" 95

2-5 Go Fujii (AIST) "Superconducting Energy-Dispersive X-ray Detector for Multi-Element Analysis

of Trace Light Elements" 97

2-6 Byeonchan Suh (NIMS) "Nanostructure Analysis of Heat Resistant Steel" 102

2-7 Taisuke Sasaki(NIMS) "3DAP/TEM study on Boron Partitioning Behavior in 9Cr-3Co-3W Heat-

Resistant Steel" 105

2-8 Kimikazu Sasa (U.Tsukuba) "6 MV Tandem Accelerator System for Ion Beam Analysis of Structural

Materials at the University of Tsukuba" 106

2-9 Akiyoshi Yamazaki (U.Tsukuba) "Development of the High Resolution Ion Microbeam System for

Analysis of Structural Materials and the Present Results of PIXE Test

Measurements at the University of Tsukuba" 110

2-10 Kenichi Kimijima (KEK) "In situ Observation of Reduction Kinetics of Iron Oxides” 113

2-11 Norimichi Watanabe (NIMS) "Characterization of Microelement in Ferritic Heat-Resistant Steels by TOF-

SIMS" 115

Theme-3

3-1 Hideaki Kitazawa (NIMS) "Overview of theme 3: Development of Multiscale Characterization in Structural

Materials" 57


5

3-2 Kazuhiro Hono (NIMS) "Microstructure Characterization of Structural Materials using Laser Assisted 3D

Atom Probe" 60

3-3 Daisuke Fujita (NIMS) "Nanoscale Characterization of Structural Composite Materials" 61

3-4 Norimichi Watanabe (NIMS) "Interface Characterization of Al/Co Laminated Film on a Si substrate using

Variable Temperature TOF-SIMS" 119

3-5 Hiroaki Mamiya (NIMS) "Precipitation induced variation of mechanical and magnetic properties for X-750

superalloy" 122

3-6 Hongxin Wang (NIMS) "Advanced In Situ Multi-functional Characterization of High Strength CFRP

Materials" 124

3-7 Tokuji Kizuka (U.Tsukuba) "3000 K Class Heating Action of Pulsed Electric Currents for High-Resolution

Transmission Electron Microscopy of High Melting Point Metals" 126

3-8 Tomoo Terasawa (U.Tsukuba) "Development of 1300 K Class Heating Stages for Transmission Electron

Microscopy of Thin Films" 129

Theme-4

4-1 Nagayasu Oshima (AIST) "Overview of theme 4: Positron Annihilation Spectroscopy based Research for

the SIP-IMASM project" 66

4-2 Akira Uedono (U.Tsukuba) "Vacancy-Type Defects and Open Spaces in Solid-State Materials Probed by

Means of Positron Annihilation" 68

4-3 Ken Wada (KEK) "Performance Test of a Pulse Stretch System for Materials Sciences at KEK Slow

Positron Facility" 74

4-4 ORourke Brian (AIST) "The AIST High-Intensity, Slow Positron Facility" 133

4-5 Lixian Jiang (AIST) "Investigation of near Surface Defects Induced by Electrical Discharge

Machining in Steel by Positron Annihilation Spectroscopy" 137

4-6 Yoshihisa Harada (AIST) "Life Prediction based on Damage for Stainless Steel" 143

4-7 Shoji Ishibashi (AIST) "Two-Component Density Functional Study of Positron-monovacancy

Interaction in Metals" 147

4-8 Shuhei Kozu (U.Tsukuba) "Evaluation of Damage by Laser Cutting Process in CFRP" 150

4-9 Gen Nakama (U.Tsukuba) "Failure and Damage Evaluation of Laser-joint in CFRP-Stainless Steel” 153

Lab Tour AIST, NIMS, University of Tsukuba, KEK）

Details to be announced.


6

Welcome to SIP-IMASM 2015

September 29th – October 1st, Tsukuba, Japan

The 1st Symposium of Strategic Innovation Promotion Program − Innovative Measurement and Analysis for

Structural Materials (SIP-IMASM) will take place from September 29 to October 1, 2015 at the AIST lecture hall,

Tsukuba, Japan. It will consist of a two day workshop and a one day lab tour. The SIP-IMASM is a R&D consortium

consisting of the National Institute of Advanced Industrial Science and Technology (AIST), the National Institute for

Materials Science (NIMS), Tsukuba University, and the High Energy Accelerator Research Organization (KEK) and

focuses on investigations of structural materials relating to aircraft and power plants. The SIP-IMASM team in a

domain of Materials Integration under the program of Structural Materials for Innovation whose program director is

professor Teruo Kishi and is working to develop unconventional measurement instruments and measurement protocols

to acquire information that is inherent in structural materials and essential for the improvement of mechanical

performance and the prediction of the lifetimes.

Our R&D resources include synchrotron radiation, ion beam analysis with a superconducting X-ray analyzer, nano-

characterization techniques such as the 3D atom probe, and positron annihilation. The SIP-IMASM team is pursuing

two strategic challenges: acquisition of “latent informative data” that have never been measured by conventional

techniques, and realization of “latent informative data” that may govern mechanical properties. Such hidden informative

data falls into four categories: “Theme-1, stress and cracks,” “Theme-2, trace light elements,” “Theme-3,

heterogeneous boundaries,” and “Theme-4, vacancies and defects,” for which four IMASM groups are responsible.

We are utilizing R&D resources in Tsukuba: three large-scale facilities such as the accelerator-based positron beam

(AIST), and high energy ion beams (TU), and synchrotron radiation (KEK). Furthermore, those facilities are being

enhanced by the use of superconducting X-ray spectroscopy detectors, a three-dimensional atom probe, and other nano-

characterization instruments such as Moire interferometry, TOF-SIMS, SPM, TEM as well as additional innovative

techniques operating at high temperature under mechanical test conditions. We are attempting to deduce hidden

informative data by forming an additional supervising group (Theme-0) for integrated analysis in addition to the four

groups. Integrated analysis is the most challenging task, which will lead to the establishment of a next-generation

innovative platform for structural materials.

In this symposium, we have invited authorities in related fields and shall present our latest R&D results in an

attempt to promote cooperation with researchers over an extensive range of measurement and analysis as well as the

above-mentioned analytical techniques.

September 29, 2015

M. Ohkubo, AIST, Chair


7

Keynote 1

Teruo Kishi


8

Structural Materials for Innovation

Teruo Kishi The Cabinet Office, Japan

Abstract: Structural Materials for Innovation (SM4I) is one of the ten R&D subjects of Cross-ministerial Strategic Innovation Promotion Program (SIPs). Three domains for materials development are assigned to (A) polymers and FRP (B) heat resistant alloys and intermetallic compounds (C) ceramics coatings. One domain called (D) materials integration performs integration of theories, experiments, computation, and measurement data.

Cross-ministerial Strategic Innovation Promotion Program (SIP) was established by the Council for Science,

Technology and Innovation (CSTI) of the Cabinet Office (CAO) in order to realize scientific and technological innovation strategically under its initiative in 2014. We emphasize industry-academia-government collaboration to establish a link between fundamental scientific research and practical materials.

Structural Materials for Innovation (SM4I) is one of the ten R&D SIP subjects [1]. Material industry of Japan, especially structural materials, forms the backbone of the whole Japanese industry. However, in addition to the United States and Europe, several emerging countries are playing catch-up with the developed countries. Therefore, strengthening the global competitiveness is one of the most important issues of Japan. Moreover, the reduction of greenhouse gas emission is also a critical issue from the viewpoint of energy and environment.

The SM4I project is aimed at the R&D goals: to produce strong, light, and heat-resistant materials for the application in transportation industry including aircrafts and energy industry, and thus to improve energy conversion and usage efficiencies. Furthermore, it is expected that the SM4I contributes to materials technologies for the aircraft and associated industries.

For the achievement of the above goals, we establish the following R&D domains on the development of aircraft engines and airframes.

(A) polymers and FRP (B) heat resistant alloys and intermetallic compounds (C) ceramics coatings (D) materials integration In addition to the R&D programs, the SM4I pursues the establishment of research centers and researcher networks

for structural materials, capacity building, and international collaboration.

Reference [1] http://www.jst.go.jp/pdf/pamph_sip_en.pdf.

Fig. 1. Framework for structural materials research supported by the government of Japan


9

Introduction to SIP-IMASM

M. Ohkubo National Institute of Advanced Science and Technology (AIST)

Department of Electronics and Manufacturing 1-1-1, Umezono, Tsukuba, Ibaraki, 305-8568, Japan

Abstract: The team of the Cross-ministerial Strategic Innovation Promotion Program - Innovative Measurement and Analysis for Structural Materials (SIP-IMASM) was formed on 26 September in 2014 in order to extend conventional analysis techniques and obtain informative data or parameters, which have never been measured before, for daccelerating the development of innovative structural materials. The first symposium is being held to promote international collaboration and report the latest results from SIP-IMASM members from September 29 – October 1 in 2015. At the meeting, we will target such structural materials as fibre reinforced plastics (FRP), heat resistant alloys, ceramic environmental barrier coating, and the other latest structural materials. The extended areas that cannot normally be covered by conventional analytical techniques include the detection of changes in the pre-deterioration stages of materials, morphological imaging at an atomic scale, and the distribution and chemical state measurement of trace light elements in addition to multiscale 2D and 3D imaging techniques. In particular, we shall focus on the chemical states of elements in texture as well as correlation between the mechanical properties of creep, fatigue, fracture, and cracks, and segregation or precipitation, deformation, and defects. The chemical states that are characteristic of the electronic structure of materials can drastically affect mechanical properties. Our innovative instruments are expected to reveal heretofore hidden information and to reveal what hidden informative data are important for further the development of innovative structural materials, and finally to contribute to the prediction of performance and lifetime for materials integration.

1. Introduction The SIP-IMASM stands for the Cross-ministerial Strategic Innovation Promotion Program - Innovative Measurement and Analysis for Structural Materials, and takes the form of a R&D consortium consisting of the National Institute of Advanced Industrial Science and Technology (AIST), the National Institute for Materials Science (NIMS), the Tsukuba University (TU), and the High Energy Accelerator Research Organization (KEK). The SIP-IMASM [1] is supported by Structural Materials for Innovation program (SIP-SM4I) [2] whose program director is professor Teruo Kishi and which is one of the ten SIPs operated by the cabinet office [3].

We intend to accelerate investigations of structural materials relating to aircraft and power plants. The above three research institutes and a university are the core members of Tsukuba Innovation Arena for Nanotechnology (TIA-nano) [4]. The SIP-IMASM is listed as a new TIA-nano project [5]. The SIP-IMASM team belongs to the domain of materials integration under the SM4I, and is working to develop unconventional analytical instruments and measurement protocols to acquire informative data that are inherent in structural materials and essential for the improvement of mechanical performance and associated lifetimes.

Synchrotron Radiation

3D Atom Probe & Nanocharacterization Ion Beam Analysis

Innovative Measurement and

Analysis forPositron Probe &

Superconducting X-ray analysis 1. Stress & Cracks

2. Trace Light Elements

3. Heterogeneous Boundaries

4. Vacancy Defects

SIP-IMASM

Fig. 1. Innovative measurement and analysis instruments (IMASM) and the four measurement areas.


10

2. Innovative measurement and analysis for structural materials (IMASM) The SIP-IMASM team is pursuing two strategic challenges: the acquisition of “latent informative data” that have never been measured by conventional techniques, and deduction of “latent informative data” that may govern mechanical properties. This hidden informative data fall into four categories: (1) stress and cracks, (2) trace light elements, (3) heterogeneous boundaries, and (4) vacancy defects, for which four IMASM groups are responsible. Figure 1 shows the four categories of the informative data and our major analytical techniques. We are employing R&D resources in Tsukuba: three large-scale facilities such as the accelerator-based positron beam (AIST), and the ion beam (TU), and the synchrotron radiation (KEK). Furthermore, those facilities have been enriched by superconducting X-ray spectroscopy, three-dimensional atom probe, and other nano-characterization instruments such as Moiré interferometry, TOF-SIMS, SPM, TEM that operate at high temperature under mechanical test condition.

We are attempting to deduce and integrate hidden informative data by forming an additional supervising group for integrated analysis in addition to the four groups. Integrated analysis is the most challenging task, which will lead to the establishment of a next-generation innovative platform for structural materials.

3. Extension of analytical techniques for structural materials Conventionally, morphological observation after crack formation or fracture is employed to analyze texture and discern the causes of the fracture or failure. We plan to extend the investigative period to the pre-deterioration stage before clear signs of deterioration such as crack formation on length scales of nm to atomic dimensions, as shown in Fig. 2. Moreover, we have added a new third axis of chemical state analysis. The chemical states that are characteristic of electronic structures determines mechanical properties. A typical example is the large difference in mechanical properties between metals and ceramics. In addition, the effects of additive elements vary with chemical state and microstructure. It is known that chemical bonding between solvent elements and trace light elements hardens metals . On the other hand, when a light element forms nano-precipitates of carbides or nitrides in metals, creep lifetime is drastically extended.

The extension area includes quantitative analysis of lattice defects or vacancy defects in the pre-deterioration stage with positron annihilation, atomic scale imaging with elemental identification with 3D atom probe, imaging of nano-scale cracks with SR-based X-ray computed tomography, nano-scale deformation imaging with Moiré interferometry, and atomic arrangement and electronic state around trace additive light elements with X-ray absorption fine structure.

4. Time schedule Analytical instruments for nanotechnology will be modified to analyze structural materials or newly developed. Some of the IMASM instruments are already running. The instruments under development are planned to be in service at the times indicated in Fig. 3.

Fig. 2. Extension concept of analytical techniques.


11

5. List of samples The five year program of the SIP-IMASM started last year. Our advanced instruments in commission for nanotechnology have already applied to structural materials and obtained numerous results that appear in these proceedings. Typical specimens of the structural materials that we are handling are listed in Table 1 with some preliminary results. Some of the preliminary experiments are beginning to acquire latent informative data that have never been obtained before.

For example, positron annihilation revealed that open space (or free volume) in polymer matrices for CFRP meaningfully varies from 0.25 nm to 0.26 nm depending on process parameters, which affect mechanical properties. Another example is an analysis of the additive light elements of nitrogen and boron in heat-resistant steels. The light elements at about 100 mass ppm are effective to increase creep lifetime by a factor of 100, but the mechanism is under discussion. We succeeded in obtaining X-ray absorption fine structure (XAFS) spectra of the nitrogen atoms in steel for the first time. The XAFS spectra are used to estimate the ratio between the nitrogen incorporated in precipitates and solid solution in the steel matrix. The measurement of boron is very difficult, but promising initial results have been achieved.

Table 1. Structural materials for SIP-IMASM and preliminary results.

SIP- domains

Materials Instruments Results

(A) FRP Synthetic resin matrix

Positron annihilation Free volume varies from 0.25 to 0.26 nm, which may affect mechanical properties.

(A) FRP CFRP Synchrotron radiation X-ray computer tomography

3D images were obtained at a spatial resolution of < 0.7 µm.

(B) Heat-resistant alloy

9Cr ferritic heat-resistant steel

X-ray absorption fine structure with superconducting detectors, 3D atom probe, TOF-SIMS

XAFS spectra of nitrogen at 71 mass ppm in steel were obtained for the first time. A ratio between nitrogen in nitrides and nitrogen in matrix may be estimated.

(D) Materials integration

Model specimen of heat-affected zone (HAZ) of steel welds

Positron annihilation, Fatigue testing machine

Vacancy defect measurement is planned.

(B) Heat-resistant alloy

Ti alloy Moire interferometry Deformation imaging of lamella phase is planned at a nm-scale spatial resolution.

(C) Ceramics

YSZ/Ni-superalloy Synchrotron Radiation X-ray computer tomography

3D images were obtained at a spatial resolution of 0.5 µm.

Fig. 3. Time schedule of SIP-IMASM.


12

coating tomography

6. Conclusions In the first symposium, we have invited J. Almer (ANL), M. Ionescu (ANSTO) from overseas, Eizo Matsumoto (Cabinet Office), Teruo Kishi (Cabinet Office), Yutaka Kagawa (Tokyo University) from the operating executive board of the SM4I project, S. Kitajima, N. Takeda, T. Kitajima, M. Enoki from the four domains of the SM4I, and K. Fukutani, M. Ohnuma, Y. Shirai, M. Kimura who are experts for the four informative needs.

We hope that our unconventional approach may initiate a new trend of the development of innovative structural materials that have a high heat resistance, long creep lifetime, high specific strength etc., and contribute to shorten a period of structural materials development by a factor of 10.

Acknowledgement This work was supported by the Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.

References [1] https://staff.aist.go.jp/m.ohkubo/SIP-IMASM/.

[2] https://staff.aist.go.jp/m.ohkubo/SIP-IMASM/SIP_materials/jst_pamphlet_english.pdf.

[3] http://www8.cao.go.jp/cstp/gaiyo/sip/sipkenkyukaihatu10kadai.pdf.

[4] https://www.tia-nano.jp/en/index.html.

[5] https://www.tia-nano.jp/en/projects/index.html.


13

Invited talk

Jonathan Almer


14

High-Energy X-ray Studies of Structural Materials at the Advanced Photon Source

Jonathan Almer1), Peter Kenesei 2), Jun-Sang-Park2), Meimei Li3), Katherine Faber4), Paul Shade5, T.J. Turner5) 1) X-ray Science Division, Argonne National Lab, 9700 S. Cass Ave, Building 431, Argonne, IL, 60439, USA. 630-252-1049 (ph),

[email protected] 2) X-ray Science Division, Argonne National Lab, IL, USA

3) Nuclear Science Division, Argonne National Lab, IL, USA 4) Dept. Of Applied Physics and Materials Science, Caltech, Pasadena, CA, USA

5) Air Force Research Laboratory, Wright-Patterson, OH, USA Abstract: A deeper understanding of how microstructure affects material properties requires information on evolving 3D structures. At present, however, the majority of such information is based on static pictures and 2D data. The reliability of structural components such as batteries, nuclear components and turbine blades are often limited by fatigue, fracture, and creep failure, in ways that are complex, and beyond simple textbook models. Combining in-situ x-ray imaging methods with simulations of microstructures and their dynamics promises to elevate our understanding of microstructure-property relationships in such systems.

High-energy x-rays from 3rd generation synchrotron sources, including the APS, possess a unique combination of high penetration power with high spatial, reciprocal space, and temporal resolution. These characteristics can be used to image microstructure with both traditional radiography and scattering modalities under a variety of environments. Over the past decade, the X-ray Science Division at the APS has developed specialized programs for these purposes, namely (i) absorption-based tomography, (ii) high-energy diffraction microscopy (HEDM), in which grain and sub-grain volumes are mapped in polycrystalline aggregates, and (iii) combined small-and wide-angle x-ray scattering (SAXS/WAXS) which permits information over a broad range of length scales to be collected from the same volume.

Applications of these techniques to study structural materials are presented. These include in situ studies- under thermal, mechanical and/or thermo-mechanical deformation - of aerospace metals [1] and high-temperature coatings [2,3], and nuclear-relevant materials [4]. Specialized equipment to conduct these in-situ studies will be presented. Efforts to process and utilize the high volume x-ray data to test predictive simulations of materials are detailed. Finally, opportunities to extend these capabilities through a proposed upgrade of the APS are discussed.

Fig. 1. Structure of a Ti-Al tensile specimen under applied load as measured with various imaging modalities: (a) near-field HEDM image colored by grain orientation, including a zoomed region of 3 grains to convey the ~2um resolution of the reconstructed boundaries, (b) far-field HEDM image showing co-axiality angle between the c-axis and loading direction for each grain and (c) absorption contrast image showing absence of voids or cracks.

From reference [1].


15

Acknowledgement Use of the Advanced Photon Source is supported by the Department of Energy, Office of Basic Energy Sciences, under contract DE-AC02-06CH11357.

References [1] J.C. Schuren, P.A. Shade, J.V. Bernier, S.F. Li, B. Blank, J. Lind, P. Kenesei, U. Lienert, R.M. Suter, T.J. Turner,

D.M. Dimiduk and J. Almer, “New Opportunities for quantitative tracking of polycrystal responses in three dimensions”, Curr. Op. Sol. State and Mat. Sci, 19 (4), 235-244, (2015).

[2] C.M. Weyant, J. Almer, KT Faber, “Through-thickness determination of phase composition and residual stresses in thermal barrier coatings using high-energy X-rays”, Acta Mat. 58 (3), 943-951 (2010).

[3] K. Knipe, A. Manero, S.F. Siddiqui, C. Meid, J. Wischek, J. Okasinski, J. Almer, A.M. Karlsson, M. Bartsch and S. Raghavan, “Strain response of thermal barrier coatings captured under extreme engine environments through synchrotron X-ray diffraction”, Nature Communications 5 (2014).

[4] L. Wang, M. Li and J. Almer, “In situ characterization of Grade 92 steel during tensile deformation using concurrent high energy X-ray diffraction and small-angle X-ray scattering”, J. Nuclear Materials 440 (1), 81-90 (2013).


16

In situ observation using synchrotron radiation of various reactions relating with steel processes

Masao KIMURA1) 2) 1) Photon Factory, Institute of Materials Structure Science, High Energy Accelerator Research Organization,

Tsukuba, Ibaraki 305-0801, Japan 2) Dept. Mater. Structure Sci., School of High Energy Accelerator Sci., SOKENDAI (The Graduate University for Advanced Studies)

Tsukuba, Ibaraki 305-0801, Japan

Abstract: As a means of in situ observation of various processes/reactions, in situ observation using synchrotron radiation, a very powerful and high-grade X-ray source, has been widely used for various fields of science because of its availability in various type of rations. We have developed in situ observation techniques for various reactions relating with steel process through cooperative research with KEK and Nippon Steel (now Nippon Steel and Sumitomo Metal). In this presentation, some typical and important results of them are shown: (a) corrosion (liquid-solid reaction), (b) sintering (non-equilibrium high-temperature reaction), (c) redox (a-short-time reaction). For further development of steel processes, new aspects in analytical approach are required: (1) revealing heterogeneity in chemical states as well as microstructure, and (2) time-resolve observation of “one way” reaction, which will be of great importance in characterization of structure materials in the field of not only steel but also aerospace. Some of these now on-going projects are also shown in the presentation.

1. Introduction

Importance of in situ observation is well symbolized by “materials science tetrahedron (MST) concept” that was originally reported by COSMAT (Committee on the Survey of Materials Science and Engineering) in 1974, chaired by Morris Cohen, MIT, and William Baker, Bell Labs. The performance of material is determined by the combination of intrinsic properties and the process that produced it, and the structure of material determines or governs these three factors and their relationship. Later in 1988, Masao Doyama, Tokyo Univ., added another important aspect: environment to MST. The materials science “octahedron” (MSO) has been the basic and still-valuable concept to research and develop a new material. In order to obtain deep understanding the reaction mechanism in terms of

MSO concept, we need to perform in situ observation of change of structure during the process for its production and/or at the time when the material is in service and really showing properties.

In order to perform in situ observation, a detecting beam must be reached into the specimen under various conditions such as in gas or water at high temperature. Furthermore, analytical technique with a high S/N ratio is required for in situ observation with a short time resolution, i.e. real-time or dynamic observation. This is why an X-ray beam and/or synchrotron radiation has been widely used for in situ observation. When a bunch of electrons is accelerated at the speed of light, it emits very intense X-ray, which is called synchrotron radiation. Synchrotron radiation is extremely bright and naturally collimated X-ray beams with a broad spectral (or energy) range (Fig. 2). The most appealing feature of observations that use synchrotron radiation is that the structural changes of a material can be observed in situ and dynamically under conditions that resemble the process or environment of interest. In this presentation, some typical and important example of in situ observation using synchrotron radiation are shown: (a) corrosion (liquid-solid reaction), (b) sintering (non-equilibrium high-temperature reaction), (c) redox (a-short-time reaction). In all cases, it is difficult to “quench” of the state during reactions, and in situ observation is the only way to reveal the reaction mechanism. And these phenomena are fundamental and important in other structural materials such as for aerospace. 2. In situ observation of atmospheric corrosion: liquid-solid reaction

Many important phenomena occur at the interface between liquids and solids in materials, such as corrosion, electrochemistry and catalytic reactions. It is difficult to quench a reaction at a liquid/solid interface, so clarifying the mechanism of the reaction requires in situ and dynamic observation of the reaction as it progresses. In order to reveal the mechanism of the corrosion of steel, we have carried out in situ observation of reactions at liquid/solid interfaces using synchrotron radiation from both “liquid” and “solid” sides of the interface.

Environment

Property

Structure

Performance

Process

In situ observation by synchrotron radiation

Fig. 1 Materials science “octahedron” (MSO) and in situ observation.


17

Fig. 2 Characteristics of synchrotron radiation and its application in steel and related fields. We performed in situ and dynamic observation of the states of chemical species near liquid/solid interfaces under the conditions in which the diffusion and electrochemical reactions ordinarily take place, using X-ray Absorption Fine Structure (XAFS) technique1), based on synchrotron radiation. The XAFS technique is a structural analysis method based on the principle that a minute change in X-ray absorption near a particular energy reflects the local atomic arrangements around the atom absorbing the X-ray. In XAFS, long-range order is not required for the specimen, and structural information is obtained from a measured spectrum by a simple and direct analysis.

The impact of this approach is well demonstrated by its application to the low-alloy corrosion-resistant steel otherwise known as weathering steel2), in which the addition of 1 mass% or less of Cu/Cr to the steel improves its weathering performance (i.e. corrosion resistance in open air). Previously, the protective mechanism was widely believed to be the formation of a dense layer of rust on the surface over several years of weather exposure, which limited the penetration of oxygen and water into the metal3,4).

However, it was not previously understood why the addition of trace elements leads to a denser layer of rust. Using various techniques including synchrotron radiation, we could directly observe the reaction in which the colloidal rust that forms in the early stages of corrosion becomes crystalline rust (crystal size: several nm) after drying. Then, from the change in atomic configuration in the rust, we could reveal the effect of additional elements in detail by our in situ and dynamic observations using synchrotron radiation5,6) 7). Based on these observations, we proposed “various scale analyses to create functioning corrosion products” 8) (Fig. 3).

Fig. 3 Schematic diagram of nano-scale reactions during atmospheric corrosion.

reaction

ΔG

in situ & real time observation in real environments

structure of nano-particles (precipitation, catalysis)

effects of alloying

surface structure in nm scale

solid/liquid (or melt) interface

Crystal Domain

(sub-grain,crystallite)

strain/stress

Δ

× 103～106

Stage I Stage II

0.5 nm 0.5 nm

0.5 nm 0.5 nm

Incoherent GB

voids

Rusts

20 nm

20 nm

Stage III

Fe

O

Cr,Cu

Fe O

grain boundary

Cu Fe

O

0.5 nm 0.5 nm Ni 20 nm

Positively-charged

Negatively-charged

Advanced weathering steel : protective rusts with fine grains and ion-selectiveness

Conventional weathering steel : protective rusts with fine grains

Mild steel γ-FeOOH α-FeOOH

γ-FeOOH α-FeOOH

γ-FeOOH

Fe2NiO4

γ-FeOOH

Fe2NiO4

Stages of atmospheric corrosion

Coherent GB


18

3. In situ observation of sintering: non-equilibrium high-temperature reaction

Iron-ore sinter constitutes the major component of the iron-bearing burden in the blast furnace in most countries in the Asia-Pacific region. Therefore, its quality and consistency have a significant impact on blast furnace performance. Iron-ore sinters is a kind of porous composite-material composed of mainly Fe2O3 and Fe3O4 grains which are attached to each other by bonding layer of various types of calcium ferrite oxides (CF) such as CaFe2O4. It is generally accepted that the quality of iron-ore sinter is governed by its microstructure, which is formed during the sintering process, as well as the properties of individual mineral phases and the size, shape, and distribution of their grains, and mutual interaction among the mineral phases9,10). In the industrial sintering process, the fine ores are mixed with limestone flux and coke breeze, and heated by the combustion of coke breeze, resulting in heating up to temperatures of 1450–1600 K (above the eutectic temperature of CaO–Fe2O3: 1478 K) for a few minutes in the region (ca. 10−1 m) near the coke breeze. As these reactions progress under non-equilibrium conditions at high temperature during a few minutes, only few reports were reported 11).

The formation of calcium ferrites (CFs) in the CaO–Fe2O3 system was investigated by in situ and real-time observation of both (a) crystal structures by using a newly developed technique, referred to as “quick X-ray diffraction (Q-XRD),” and (b) microstructures by using an in situ laser microscope. In the new Q-XRD, a specimen was heated up to 1773 K, and X-ray diffraction patterns were measured using a pixel-array area detector with an interval as short as a few seconds (Fig. 4).

Reaction schemes for different conditions of heating and cooling rates for specimens various chemical compositions. Based on these results, he first continuous cooling transformation (CCT) concept for iron ore sintering was proposed to understand overcooling phenomena when the molten oxide cooled down to room temperature and magnetite (Fe3O4), hematite (Fe2O3), and various types of calcium ferrite were formed12) (Fig. 5). The CCT diagram for sintering provides crucial and fundamental information on the sintering accompanying solidification, precipitation, and formation of calcium ferrites from the molten oxide, and can be used as a guideline for controlling sintering processes.

Fig. 4 (a) X-ray geometry of the Q-XRD system and (b) the reaction chamber in the Q-XRD system. Fig. 5 First continuous cooling transformation (CCT) diagram for sintering of specimen CaO:Fe2O3 = 10:90 mass%12).

4. In situ observation of reduction and oxidation: a-short-time reaction

In iron-production and relating process, a various type of gas was used and/or produced. Thus a variety of catalytic reactions were widely used to make the best use of it, where the reduction and oxidation (redox) reactions is one of the most importance. Here we represent in situ observation of redox reactions in Pd/Sr-Fe-O catalyst with a time resolution as short as ms. We have developed the Pd/Sr-Fe-O catalyst which exhibits high performance for automotive emission

Δ

Time, t / s

Tem

pera

ture

, T /

K

Δ

Δ

Slow cooling Fast cooling

Area detector PILATUS ®

X-ray


19

control. The Sr-Fe-O oxide support has a unique “multi-phase-domain” structure, where a single grain is composed of nano-sized domains of three phases: SrFeO3-δ, Sr4Fe6O13-δ, and SrFe12O19-δ

12,13). A series of time-resolved Pd K-edge absorption near edge structure (XANES) spectra of the Pd/Sr-Fe-O catalyst could be successfully obtained by in situ DXAFS at ca. 700 K during the redox cycles. Time-resolved wavelength dispersive XAFS (DXAFS) spectra at the palladium K-edge over the energy range (23.8-25.3 keV) were measured on the NW2A station of the Photon Factory Advanced Ring (PF-AR) at the High Energy Accelerator Research Organization (KEK) in Japan. In DXAFS, we can obtain the structural information of a specific element without scanning energy, resulting in a time resolution as short as ms.

Detailed analysis of in situ and simultaneous observation of palladium redox and oxygen storage/release revealed a strong correlation between the redox of palladium and the oxygen storage/release in the Pd/Sr-Fe-O catalyst as follows. Reduction from PdII to Pd0 begins just after the introduction of H2/He gas, with a simultaneous increase in the oxygen deficiency (δ) in perovsikite-type SrFeO3-δ and Sr4Fe6O13-δ. After the completion of the Pd reduction, the recovery (decrease) of δ follows. Contrary to this sequence, oxidation from Pd0 to PdII begins after an incubation period, after which the change in δ precedes. The palladium oxidation and the increase of δ then progress simultaneously (Fig. 6). 12,13)

Fig. 6 Schematic illustration of the reaction mechanisms revealed by in situ observation, showing the oxidation process experience a reaction of recovery of oxygen defects of the substrate and the reaction speed is slower than the reduction.

5. Summary

Importance of in situ observation was emphasized by some typical and important results: (a) corrosion (liquid-solid reaction), (b) sintering (non-equilibrium high-temperature reaction), (c) redox (a-short-time reaction). As shown, in situ observation of information averaged over a specimen with a time resolution of s – ms have been established and widely used in may fields. For further development of materials science, understanding the reaction mechanism in terms of MSO concept, new aspects in analytical approach are required: (1) revealing heterogeneity in chemical states as well as microstructure, and (2) a very-short time-resolve observation of “one way” reaction. They are being realized using synchrotron radiation by X-ray microscopy and laser-triggered observation. In SM4I (Structural Materials for Innovation) project in SIP program, we plan to develop those characterization approaches and apply them to structure materials for aerospace.

Acknowledgement

Most parts of this work have been performed under the cooperative research between Nippon Steel & Sumitomo Metal Co. and KEK. I would like to thank all persons involved in the project.

vacuum Ox-cycle: 20%O2/He ρ

Ox-I Ox-II

Pd(I

I)/(

Pd(I

I)+

Pd(0

))

Ox-II

Oxidation of Pd

Ox-I

O Vacancies

Pd-H

Ox-0

O H in Pd Pd

ρ

Red-I Red-II

4%H2/He vacuum Red-cycle:

Pd(I

I)/(

Pd(I

I)+

Pd(0

))

Fe Pd O Sr (omitted)

Red-I

Pd(II) Pd(0)

Red-II

Relaxation of lattice


20

References [1] D. C. Koningsberger, in X-Ray Absorption:PRINCIPLES, APPLICATIONS, TECHNIQUES OF EXAFS, SEXAFS

AND XANES, edited by D. C. Koningsberger and R. Prins (John Wiley & Sons, New York, 1988). [2] C. B. Larabee and S. K. Coburn, in The Atmospheric Corrosion of Steels as Influenced by Changes in Chemical

Composition [Corten (U.S. Steel)], London, U.K., 1962, p. 276. [3] H. Kihira, S. Ito, and T. Murata, Corrosion 45, 347 (1989). [4] H. Okada, Y. Hosoi, K. Yukawa, and N. H., J. Iron Steel Inst. Japan (TETSU-TO-HAGANE) 56, 277 (1970). [5] M. Kimura, H. Kihira, N. Ohta, M. Hashimoto, and T. Senuma, Corros. Sci. 47, 2499 (2005). [6] T. Mizoguchi, Y. Ishii, T. Okada, M. Kimura, and H. Kihira, Corros. Sci., 2477 (2005). [7] M. Kimura, N. Ohta, and H. Kihira, ECS Transactions 16, 63 (2009). [8] M. Kimura, T. Mizoguchi, H. Kihira, and M. Kaneko, in Characterization of Corrosion Products on Steel

Surfaces, edited by Y. Waseda and S. Suzuki, 245-272, Chap.11 (Springer, 2006). [9] S. N. Ahsan, T. Mukherjee, and J.A.Whiteman, Ironmaking and Steelmaking 10, 54 (1983). [10] L. Lu, R. J. Holmes, and J. R. Manuel, ISIJ International 47, 349 (2007). [11] A. S. Nathan, M. I. Webster, I. Pownceby, C. Madsen, and J. A. Kimpton, Metall. Mater. Trans. B 43, 1344 (2012). [12] M. Kimura and R. Murao, ISIJ International 53, 2047 (2013). [13] M. Kimura, K. Uemura, T. Nagai, Y. Niwa, Y. Inada, and M. Nomura, J Phys. Conference Series, Vol.190,

012163 (2009).


21

Development of Ceramic Environmental Barrier Coatings for Advanced Airplane Engine Applications

Satoshi KITAOKA 2-4-1 Mutsuno, Atsuta-ku, Nagoya 456-8587, Japan,

Phone: +81-52-871-3500, Fax: +81-52-871-3599, E-mail: [email protected]

Abstract: Environmental barrier coatings (EBCs) can play a key role in allowing SiC fiber reinforced SiC ceramic matrix composites (SiC/SiC) to be applied to advanced hot section components for better environmental stability and thermomechanical durability. The advanced EBC, which maximizes the performance of SiC fibers excellent in heat resistance, applicable under water vapor environments at high temperatures is being developed in the SIP project. The EBC is multilayered, consisting of a bond layer, an oxygen shielding layer, a water vapor shielding/volatilization barrier layer, and a relaxation layer against thermal shock, in order from the SiC/SiC substrate side to the top coat. Mass transfer in the EBC should be effectively controlled so as to enhance the shielding properties and high-temperature stability of the multilayered structure. The importance of the design of EBCs based on the the mass-transfer mechanisms within the EBC constituent materials is explained using the example of alumina scale on heat-resistant alloys. 1. Introduction Improvements in the heat resistance of airplane jet engine components, as well as weight reductions, are essential in order to reduce both the fuel consumption and CO2 emissions of these engines. There is a particular need for structural materials that may be used in hot section components and that are highly durable when exposed to severe environments, including water vapor at high temperatures. The limiting temperature for conventional Ni-based superalloys is about 1100 °C. Even with further improvements in current alloys and incorporation of air cooling, the limiting temperature is merely increased to around 1200 °C. In recent years, there has been an interest in using SiC fiber-reinforced SiC matrix composites (SiC/SiC) for these applications, because these materials are extremely light weight and offer superior heat resistance compared to the alloys. In particular, the most heat resistance SiC fibers, which are fabricated in Japan, have a practical limiting temperature of about 1400 °C. However, it is well-known that SiC/SiC deteriorates in water vapor environments above approximately 1100 °C due to oxidation and subsequent volatilization. EBCs can play a key role in allowing SiC/SiC to be applied to advanced high-pressure turbine components with improved environmental stability and durability. Nowadays, the limiting temperature for EBCs, at which they can find commercial application, is about 1300 °C. Therefore, an increase in the limiting temperature of EBCs will enable SiC/SiC to be applied under severe conditions such as higher temperatures for a long time.

EBCs must have excellent environmental shielding and thermomechanical durability in addition to excellent volatilization resistance. In the SIP project, a multilayered structure is applied in the case of the EBC under development with the goal of achieving exceptional performance by the overall EBC system through the use of layers with individual characteristic functions. Fig. 1 shows architecture of advanced EBC system. The EBC includes a bonding layer on the SiC/SiC substrate, followed by dense layers including an oxygen shielding layer and a water vapor shielding/volatilization barrier layer. The EBC is made using a segmented structure to reduce thermal stresses during temperature cycling, thus producing a thermal barrier coating (TBC). The gradient change in the Yb silicate layer composition from the disilicate to the monosilicate enhances the volatilization barrier function, and simultaneously increases compatibility with the underling layer.

Mass transfer in the EBC should be effectively controlled so as to enhance the shielding properties and high-temperature stability of the multilayered structure. Therefore, the importance of the design of EBCs based on the mass-transfer mechanisms within the EBC constituent materials will be explained using the example of α-alumina scale on heat-resistant alloys.

2. Control of mass transfer in α-alumina Ni-based and Fe-based superalloys become oxidized by exposure to high operating temperatures above 900 °C, resulting in eventually formation of polycrystalline α-alumina scale. The durability of the alloy components in such oxidizing ambient strongly depends on the barrier’s performance with respect to oxygen permeation through the scale. The scale growth during the oxidation is regulated by diffusion of oxygen and aluminum ions along grain boundaries (GBs) in the alumina layer in response to their respective chemical potential gradients. Small quantities of oxygen-reactive elements, such as Y, Zr and Hf, are often added to the alloys in order to improve their oxidation resistance. These elements segregate at GBs in growing alumina scales during the oxidation process, and are thought to suppress

Fig.1 Architecture of advanced EBC system.


22

mass transfer along the GBs, as well as to improve scale adhesion under thermal cycling conditions. This study focuses on control of oxygen and aluminum GB diffusion in alumina by such segregated dopants.

The oxygen permeability technique using polycrystalline α-alumina wafers is expected to be useful for investigating the mass transfer mechanisms in the scales. These wafers, with a thickness of several hundred microns, were exposed to an oxygen potential gradient (ΔμO) at high temperature, with the two surfaces of the wafer deliberately subjected to different oxygen partial pressures (PO2(hi)>>PO2(lo)).[1,2] Because high purity polycrystalline alumina has an excellent oxygen shielding property, the oxygen permeability measurements using a zirconia oxygen sensor must be carried out at temperatures above 1500 °C so as to accelerate mass-transfer in the alumina wafers, aiding the detection of a tiny amount of oxygen molecules permeated through the wafers. At such high temperatures, oxygen permeation occurred by GB diffusion of oxygen from the PO2(hi) surface to the PO2(lo) surface, along with simultaneous GB diffusion of aluminum in the opposite direction, similar to the case for mass transfer through alumina scales. It was found that the oxygen and aluminum fluxes at the outflow side of the wafer were significantly larger than those at the inflow side [1]. The narrow bandgap of incoherent GB suggests to be responsible for the switching behavior of the main diffusion species [3]. The effect of GB segregation of Hf or the rare-earth elements (Ln) such as Y and Lu on mass transport has also been previously investigated for a large ΔμO at temperatures above 1500 °C [1]. Very interestingly, Hf and the rare-earth elements were found to selectively reduce the diffusivity of aluminum and oxygen, respectively. The reduction in oxygen diffusivity was found to be similar for Y and Lu. In all cases, the reduction in diffusivity was not associated with a change in the activation energy for diffusion, given by the Arrhenius equation, but was instead due to a change in the frequency factor.

In the present study, to further improve the oxygen shielding capability of alumina layers, the effect of Hf and Ln dopants on GB diffusion of aluminum and oxygen was investigated for alumina wafers exposed to a large ΔμO at high temperature. While past research was limited to the effect of a single type of dopant, the present study also considers a Hf and Ln co-doped single wafer, and a bilayer wafer containing adjacent Hf- and Lu-doped layers exposed to different PO2 gradients. Figure 2 shows the calculated oxygen and aluminum flux profiles in different wafers exposed to PO2(hi) and PO2(lo) partial pressures of 105 and 10-8 Pa, respectively, at a temperature of 1600 °C. The calculation procedure has been described in detail elsewhere [2]. The dotted lines in the figure indicate the summation of these fluxes and corresponds to the oxygen permeation in the steady state. In the case of undoped wafer, it is worth noting that both fluxes at the outflow side are significantly larger than those at the inflow side, in accordance with dominant Al transport at the PO2 (hi) side and dominant oxygen transport at the PO2 (lo) side. In the case of the bilayer wafer, in which a Y-doped layer is exposed to the lower PO2 side and an Hf-doped layer is exposed to the higher PO2 side and each layer has the same thickness, the sum of both fluxes is obviously decreased. However, when the bilayer structure is reversed, the maximum fluxes are similar to those in the undoped wafer. The calculated oxygen permeability constants were found to be in good agreement with the experimental values obtained from the oxygen permeation tests. On the other hand, for the co-doped wafer, the oxygen permeability (data not shown) was higher than that for the undoped wafer because of the formation of Ln-stabilized HfO2 particles at the GBs, which acted as extremely fast oxygen diffusion paths. These results demonstrate that a suitable choice of dopant configuration, taking into account the nature of the diffusing species and the role of the dopants, allows optimization of the high-temperature oxygen shielding capability of alumina. 3. Expectations for IMASM In the SIP project, the design and usage limit prediction of the advanced EBC will be also based on the mass transfer mechanisms within the EBC exposed to a large ΔμO similar to that applied in alumina protective film. The EBC is a multilayered structure that is composed of some complex oxides so that the mass transfer mechanisms within the EBC is considered to be much more complicated than those of alumina. Therefore, our goal is to develop advanced coating technologies offering both excellent environmental shielding and thermomechanical durability based on ongoing coordination between the efforts of our ceramic coating group and the IMASM center. References

[1] T. Matsudaira et al., J. Am. Ceram. Soc. 96 (2013) 3243.

[2] S. Kitaoka et al., J. Am. Ceram. Soc. 97 (2014) 2314.

[3] T. Ogawa et al., Acta Materialia 69 (2014) 365.

Fig.2 Oxygen and aluminum flux profiles in alumina wafers exposed to PO2(hi) and PO2(lo) partial pressures of 105 and 10-8 Pa, respectively, at a temperature of 1873 K.


23

Overview of theme 1: Imaging of strain, cracks and chemical states for structural materials

Masao Kimura1,2), Yasuo Takeich1,2), Yasuhiro Niwa1), Yasuhiro Niwa 1), Kenichi Kimijima1), Hiroaki Nitani1,2), Yumiko Takahashi1), Keiichi Hirano 1,2), Kazuyuki Hyodo1,2), Yoshihisa Tanaka3), Qinghua Wang4), Shien Ri4)

1) Photon Factory, Institute of Materials Structure Science, High Energy Accelerator Research Organization, Tsukuba, Ibaraki 305-0801, Japan

2) Dept. Mater. Structure Sci., School of High Energy Accelerator Sci., SOKENDAI (The Graduate University for Advanced Studies) Tsukuba, Ibaraki 305-0801, Japan

3) National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan 4) Res. Institute for Measurement and Analytical Instrumentation, National Institute of Advanced Industrial Science and Technology,

Tsukuba, Ibaraki 305-8568, Japan

Abstract: Recently, microstructures of structural materials, such as ceramic-coating, steel, nickel-alloy, FRP (fiber-reinforced polymer), have been developed to have much more complicated three-dimensional (micro-) structure composed of multi-phases in a wide range from nm to mm, and used in conditions where the distribution of chemical potential and stress is wide and changing during in service. Thus, characterization of heterogeneity in those materials in the two- and three-dimension is essential to research and develop new structural materials for aerospace. In the theme 1: “stress and cracks”, we have developed multi-scale and complementary techniques for imaging of strain, cracks and chemical states for structural materials using: moiré and digital image correlation method, X-ray computed tomography (X-CT), X-ray absorption spectroscopy (XAS), X-ray diffraction (XRD). We applied them to typical structural materials, showing their feasibility and potential to reveal mechanism of degradation of them during in service under severe conditions such as at high temperatures under reaction gas with mechanical load.

1. Introduction As simply presented in the Griffith-Orowan fracture theory (Eq.1), the criteria for propagation of crack can be determined by the following factors: (a) strain and/or stress distribution around the crack, (b) the size and morphology of a crack, and (c) the energy change by the formation of “cracked surface”. As microstructures of structural materials, such as ceramic-coating, steel, nickel-alloy, FRP (fiber-reinforced polymer), have been developed to have much more complicated three-dimensional (micro-) structure composed of multi-phases in a wide range from nm to mm, and used in conditions where the distribution of chemical potential and stress is wide and changing during in service. Thus, in the theme 1: “stress and cracks”, we have planned to research and develop multi-scale and complementary techniques to characterize the heterogeneity these three factors, i.e. or realize mapping or imaging them in the two- and three-dimension in structural materials for aerospace.

(Eq.1)

As the first step, we have tried to utilize advanced techniques that we had developed in the field of physics and chemistry for mapping or imaging of (a) strain and/or stress, (b) cracks, and (c) chemical states, in relating with microstructures of structural material. In this presentation, we showed the typical results at this stage, and discuss them for achieving the next step. Detailed results are shown by other presentations. 2. Imaging of strain and/or stress Special attention has been focused on measurement method on a multi-scale deformation and strain distribution in a hierarchical microstructure composite material during mechanical and thermal loading via in situ field emission scanning electron microscope (FE-SEM) observations. Macroscopic deformation behaviors are observed by electron moiré method and digital image correlation method at the different length scales, during 3 point-flexure test. (Figs.1 and 2) .

It was shown that the proposed experimental technology can provide inhomogeneous deformation and strain behaviors at multiple length scales and their related boundary conditions such as interface debonding and sliding, damage initiation and evolution. The results obtained are expected to be used widely as effective tool for contributing the experimental mechanics and reliability engineering for advanced structual materials.


24

3. Imaging of cracks and microstructure By employing a unique X-ray optics, we have achieved X-ray computed tomography (CT) measurements with a resolution as high as 0.7 μm even using a in-house X-ray source. Figure 3 shows a typical result of CFRP, where each fibber and the adhesive layer between the bundle of fibbers were clearly observed with a resolution of ca. 0.7 μm. (a) (b) For further advanced X-ray computed tomography (X-CT) measurements, we use synchrotron radiation. When we performed CT measurements with a single energy by using monochromator, we can obtain 3D CT images with high-contrast and less artifacts caused by such as the beam hardening effect. Figure 4 shows volume rendering of silicon carbide fiber-reinforced silicon carbide matrix (SiC/SiC) by synchrotron radiation X-CT. The fiber bundle structure and the porous network were clearly observed.

4. Imaging of chemical states We have developed the new beamline BL-15A1 at Photon Factory, IMSS, KEK, Japan. X-ray absorption fine structure (XAFS) and XRD from the specimen surface in an area with ca. 20μm in diameter were measured by scanning the specimen. XAFS spectra were measured using SDD (silicon drift detector) in the fluorescence geometry (Fig. 5). Partially-reduced iron-ore sinters were analyzed as a model system to establish the chemical mapping approach. Iron-ore sinters is a kind of porous composite-material composed of mainly Fe2O3 and Fe3O4 grains which are attached to each other by bonding layer of various types of calcium ferrite oxides (CF) such as CaFe2O4. Chemical states of Fe in the specimen was determined by the ratios of fluorescence intensities measured for different energies around Fe K-α edge. Figure 6 shows a typical example of XAFS-mapping for the cross section of sinter which was in the middle of reduction process. Most parts in the specimen were in the mixture state of FeIII and FeII corresponding to Fe3O4 phase,

Fig. 1 Macroscopic deformation behaviors of CFRP observed by the electron moiré method under different applied strains of (a) εa = 0, and (b) εa = 0.005 .

Fig. 2 Another example of (a)Moiré fringes and (b) relative displacement distributions of CFRP under three-point bending (1548μm×612μm).

Fig. 3 X-CT images of CFRP for (a) low and (b) high magnification. Arrows in Fig3(b) show a kind of voids among the fibbers.

z x y

(a) (b)

Fig. 4 (a) 3D volume rendering image, and (b) porous network of the SiC/SiC. Scale bars indicate 1mm.


25

which was also confirmed by XRD measurements. However, it was clearly shown that there exist regions where the chemical state are in mainly FeII (regions A in Fig.6) and mainly FeIII (regions B), respectively.

5. Summary We have tried to utilize advanced techniques that we had developed in the field of physics and chemistry for mapping or imaging of (a) strain and/or stress, (b) cracks, and (c) chemical states, and showed their feasibility for structural material. We plan to focus on specific systems in ceramic-coating and CFRP, and apply these techniques for revealing initiation and propagation of degradation at high temperatures under reaction gas that is one the most important properties as structure materials for aerospace.

We also plan to develop a brand-new characterization technique: XAFS-CT nano-imaging microscopy, where 3D imaging of chemical states as well as microstructure can be obtained simultaneously with a resolution as high as 50 nm. We have performed preliminary test experiments and finished designing the system. Acknowledgement A part of this work was supported by Cross-ministerial Strategic Innovation Promotion Program (SIP, unit D66) operated by the cabinet office. A part of this work has been performed under the approval of the Photon Factory Program Advisory Committee (Proposal No. 2014G707 and 2015S2-002).

Fe A1

A2

B1 B2

B3

Fig.6 Chemical state mapping of Fe by XAFS in the partially-reduced

Fig.5 Outline of X-ray optics for chemical mapping at BL-15A1, KEK.


26

In-situ multi-scale strain imaging for composite materials using FE-SEM during mechanical and thermal loading

Yoshihisa Tanaka*, Kimiyoshi. Naito, Satoshi. Kishimoto National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan,

Phone and fax numbers: +81-29-859-2240 and +81-29-859-2401 *Corresponding author’s email: [email protected]

Abstract: The multi-scale pattern has been developed to characterize the full-field deformation and strain distribution at different scales by electron beam lithography ranging from nano to millimeter scales. The multi-scale pattern was applied to CFRP laminate with hierarchical structure during mechanical and thermal loading by using in-situ FE-SEM observation. The results obtained are expected to be used widely as effective tool for understanding the interaction between local and macroscale interfacial characterization and effect of the composite mechanical performance.

1. Introduction

Currently, the increasing usage Carbon Fiber Reinforced Polymer composite (CFRP) and hybrid materials in load bearing applications such as aircraft structure requires method to predict the mechanical properties and damage mechanisms, in order to design more damage tolerant structures. The composites exhibit hierarchical microstructure having highly anisotropic and heterogeneous over a wide range of the length scales and failure mechanisms can be controlled by several scales. Thus the thermomechanical properties (coefficient of thermal expansion (CTE), modulus, and strength) in the transverse of the carbon fiber are considerably different with longitudinal direction due to highly anisotropic microstructure [1-2]. The interface mechanical characterization and determination during mechanical and thermal loading has been always a field of major interest in CFRP and hybrid materials, because the interface bond strength plays an important role in determining the strength of these composite materials. The development of damage progressed differently in multidirectional interface in CFRP laminates is characterized as localized matrix plasticity at the interfaces between fiber and matrix and laminates with different length scales, and large delamination causes at the laminate interface in hybrid material due to large mismatch of CTE between CFRP and metal (Figure 1). However, the measurement method of the interfacial characterization in CFRP at different scales is considerably difficult during mechanical and thermal loading, since the measurement approaches of interfacial strength are mainly using different specimen and testing such as micromodel composite (single fiber tests, fragmentation test, etc.) and interlaminar fracture testing (the double cantilever beam (DCB) for mode I, the end notch flexure (ENF) for mode II) using macroscopic specimens [3-4].

In the present study, special attention has been focused on measurement method on a multi-scale deformation and strain distribution in a hierarchical microstructure composite material during mechanical and thermal loading via in-situ field emission scanning electron microscope (FE-SEM) observations and the effect of the local characterization on macroscale response.

2. Experimental procedure 2.1 Hierarchical composite materials

For the application of the multiscale deformation and strain measurement during mechanical and thermal loading, the material used in this study was CFRP laminates consisting of an ultrahigh strength PAN-based carbon fiber, an

Fig. 1 Failure behaviors in CFRP

Fig. 2 Microstructure of cross-ply CFRP in (a), (b) PAN based carbon fiber, and (c) pitch based carbon fiber.


27

ultrahigh modulus pitch-based carbon fiber, and an epoxy matrix because of their hierarchical internal structures over various length scales. The laminates were composed of seven plies with 0 and 90 orientations by stacking sequence, as

shown in Figure 2. The fiber volume fraction was approximately 0.6.

2.2 In-situ multiscale strain imaging method

In order to measure multiscale deformation and strain distribution during mechanical and thermal loading, In-situ FE-SEM (Quanta 200 FEG, FEI Corp.) observations of the damage evolution process during loading were carried out by using the mechanical and thermal loading devices installed into sample chamber (Figure. 3(b) and 3(c)). The mechanical load increment was controlled precisely by a stepping motor of the loading device and the thermal load was controlled by heating/cooling stage which the operating temperature is ranging from 80 to 730K. Electron beam lithography system was installed into a FE-SEM system to develop the pattern from nanoscale to macroscales (Figure. 3(a)). Figure 4 shows a multiscale pattern composed of a grid as well as random and nanocluster was developed and introduce onto the specimen surface to measure the multiscale deformation and strain distribution [5]. The grid and random patterns were fabricated by electron beam lithography and nanocluster was fabricated by sputtering method. The macroscale deformation of the specimen was measured using the grid pattern produced in an FE-SEM through the interference between the electron beams (reference grid) and the grid (master grid) by the electron moiré method [6]. The microscale and nanoscale deformations were measured and analysed by the digital image correlation (DIC) method using the random and the nanocluster patterns with before and after deformation images. The damage evolutions in CFRP during the three-point flexure loading and the thermal loading with the operating temperature range of 200K (170 – 370K) were investigated by in-situ FE-SEM observations from nanometer to millimeter scales.

3. Results and discussions Macroscopic deformation

behaviors are observed by electron moiré method and digital image correlation

Fig. 3 (a) Experimental setup for in-situ FE-SEM observation with electron beam lithography system, (b) installed mechanical loading device (maximum load: 5KN), and (c) heating/cooling stage (80 – 770K) into the chamber.

Fig.4 Typical example of a multiscale pattern consisting of (a) a grid, (b) a random pattern, and (c) a nanoscale pattern, observed by backscattered electron imaging (BSEI) at different magnifications..

Fig. 5 Macroscopic deformation behaviors observed by the electron moiré method under different applied strains of (a) εa = 0, and (b) εa = 0.005, and shear strain distribution around carbon fibers at the magnification of the boxed region in (b).


28

method at the different length scales, during 3 point-flexure test, as shown in Figure 5. The moiré fringe patterns are almost straight before deformation. With increasing bending deformation, the moiré fringes are distorted from the original fringes, and then the distorted angle has the maximum at the tip and its decreases with distance from the loading tip. Especially, delamination is clearly observed at the laminate interface, indicated by arrow in Figure 5(b). The moiré pattern was changed after the delamination at the laminate interface between 0 and 90 degree cross-ply pitch-based carbon fiber laminates caused by the shear deformation. The shear strain distribution around PAN-based carbon fibers analyzed by DIC method revel the maximum localized shear strain located at 45 degree and it strong depends on the fiber distribution (Figure 5(c)).

Figure 6 shows the macroscopic deformation observed by the electron moiré method under the different temperature of 170 and 370K for the cross section of CFRP laminate. The moiré fringe patterns were clearly generated in the region of the grid pattern with 250 nm spacing. The spacing of the moiré fringe lines were decreased with increasing temperature due to the macroscopically thermal deformation. The fringe lines in the cross section of pitch-based carbon fibers are locally distorted with increasing the temperature due to inhomogeneous deformation by inhomogeneous fiber distribution (Figure 6(b)). The inhomogeneous thermal deformation around the fibers is clearly observed near the interface and the interface debonding appears at the fiber/matrix interface, at the temperature range of 200K (Figure 6(d)). This suggests that the nano scale deformation and strain distribution in the small area will affect the interfacial bonding strength due to the anisotropy and inhomogeneous microstructure. This measurement method can be used to provide higher-order of thermal expansion behaviors, such as thermal expansion inhomogeneity and anisotropy, interface delamination, deformation gradients needed for developing the thermal damage mechanism understanding of the CFRP.

4. Conclusion The experimental technology proposed provides inhomogeneous deformation and strain behaviors at multiple length scales and their related boundary conditions such as interface debonding and sliding, damage initiation and evolution. The results obtained are expected to be used widely as effective tool for contributing the experimental mechanics and reliability engineering for advanced structual materials.

Acknowledgement A part of this research was supported by "Cross-ministerial Strategic Innovation Promotion Program (SIP), Structural Materials for Innovation" from Japan Science and Technology Agency, JST.

References [1] Chand S., “Review – carbon fibers for composites”, J. Mat. Sci., 35(6), 1303–1313 (2000).

[2] Kulkarni R. and Ochoa O., “Transverse and longitudinal CTE measurements of carbon fibers and their impact on interfacial residual stress in composites. Journal of Composite Materials”, 40(8), 733-754 (2006).

[3] X.F. Zhou, H.D. Wagner, S.R. Nutt, “Interfacial properties of polymer composites measured by push-out and fragmentation test”, Composite A, 32, 1543-1551 (2001).

[4] K. Friedrich, (Eds.) [Application of fracture mechanics to composite materials], Elsevier Science Publishing, USA, (1989).

[5] Y. Tanaka, K. Naito, S. Kishimoto and K. Kagawa, “Development of a pattern to measure multiscale deformation and strain distribution via in situ FE-SEM observations”, Nanotechnology, 22, 115704-115709 (2011).

[6] S. Kishomoto, “Electron moiré method”, J. Soc. Exp. Mech., (3)1, 9-14, (2003).

Fig. 6 Microscopic thermal deformation behaviors observed by the electron moiré method under different temperatures of (a) 170K, and (c) 370K. The relative thermal deformation around the fiber analyzed by DIC method, (a) at temperature range of ΔT = 40K, and (b) of ΔT = 200K.


29

Application of XAFS & XRD mapping techniques for various materials

Y. Takeichi1,*), H. Nitani1), R. Murao2), N. Ohta3), K. Noami3), Y. Niwa1), K. Kimijima1), and M. Kimura1) 1) Photon Factory, High Energy Accelerator Research Organization (KEK), Tsukuba, Ibaraki 305-0801, Japan

2) Advanced Technology Research Laboratories, Nippon Steel & Sumitomo Metal Co., Futtsu, Chiba 293-8511, Japan 3) Nippon Steel & Sumikin Technology Co. Ltd., Futtsu, Chiba 293-0011, Japan

*[email protected]

Abstract: An instrument for X-ray spectromicroscopy at the Photon Factory is reported. X-ray absorption, fluorescence, and diffraction patterns of the specimen are simultaneously mapped with the spatial resolution of ~20 μm. With the photon energies of 2.1–15 keV, fruitful information about chemical properties of a specific region of interest is accessible. The method is capable of various in situ/operand experiments, since it is photon-in, photon-out experiment and allows wide space left for sample environment. A result of reduction stage mapping in iron-ore sinters was presented.

1. Introduction Performance of structural materials depend not only on the bulk composition or chemical properties, but also on the "heterostructures" in the materials. Many of the materials have its natural heterostructures as seen in fibre-reinforced plastics. A deterioration of structural materials, such as cracking or chemical degradation, stems from a weakest point or reaction centre, and then propagates into bulk. Therefore, it is important to investigate the deterioration process by microscopic technique with a scale corresponding to the phenomena. Microscopies using energy-tunable X-rays, such as X-ray absorption fine structure (XAFS) imaging or X-ray fluorescence (XRF) imaging, provide fruitful information about chemical properties of a specific region of interest, and therefore are one of the promising technique to investigate the deterioration phenomena of structural materials.

We have developed a semi-μXAFS apparatus to investigate heterogeneous properties of structural materials. An X-ray photon-in photon-out technique allows simultaneous investigation with various detection methods such as transmission, fluorescence and diffraction. Moreover, it is suitable to in situ experiments by varying temperature, applying compressive/tensile force, or applying oxidation/reduction gas atmosphere. The spatial resolution of scanning X-ray microscopes is determined by a size of focused X-rays, while penetration depth of the probe depends on the photon energy; it is tens of nanometers in soft X-ray region, and is micrometers in multi-keV region. In this report, we describe the instrument and present a result of a chemical mapping experiment using it.

2. BL-15A1: a semi-μXAFS endstation The instrument is placed at a hard X-ray undulator beamline, BL-15A1 of the Photon Factory, Japan [1]. The mirror optics of the beamline provides a intense X-ray beam of ~1011 photons/sec focused down to ~20 μm. As illustrated in Fig. 1, the endstation is equipped with a ion chamber for transmission XAFS, silicon-drift detector (SDD) for XRF, and two-dimensional detector (PILATUS 100k) for X-ray diffraction (XRD) pattern measurement. Transmission XAFS, XRF, and XRD patterns can be obtained simultaneously. Therefore, this equipment is capable of mapping of elemental composition, chemical properties such as oxidation stage, and crystal structures in tens-of-μm scale. X-ray beam size here is determined by the mirror optics that realizes achromatic focusing for a wide range of photon energies, 2.1–15 keV. The beam size is evaluated in an edge-scanning manner as shown in Fig. 2.

Fluorescence

FluorescenceTransmission

Diffraction

Semi-microX-ray beam

Fig. 1. Schematic design of XAFS/XRF/XRD mapping instruments at BL-15A1 semi-μXAFS endstation.


30

(a)

Horizontal

FWHM:19.8μm

3.0

2.5

2.0

1.5

1.0

0.5

0.0

60

50

40

30

20

10

0

Differential intensity (arb. units)

Inten

sity (

arb. u

nits)

-150 -100 -50 0 50 100 150Position (μm)

(b)

Vertical

FWHM:21.3μm

3.02.52.01.51.00.50.0

15

10

5

0

Differential intensity (arb. units)

Inten

sity (

arb. u

nits)

-150 -100 -50 0 50 100 150Position (μm)

Fig. 2. Beam size at the sample position of semi-μXAFS endstation in (a) horizontal and (b) vertical directions.

The sample position is scanned with motorized linear stages. The sample and detector angles can be varied with a goniometer to observe appropriate XRD patterns and to avoid diffraction X-rays deteriorating XRF signal. The sample surface can be monitored using an on-line optical microscope. A wide space is kept for the sample environment as conventional XAFS beamlines. The sample stage can hold a specimen with the size up to centimeters, or cells for in situ/operando measurements.

3. Chemical mapping of iron-ore sinters Steel is one of the most common structural materials because of its high tensile strengths, richness in natural resources and low costs. Iron-ore sinters are the major component of a blast furnace iron-bearing burden. The performance of the blast furnace depends on the quality of the iron-ore sinters that have micrometer-scale heterostructures; they consist of iron-oxide mineral phases (α-Fe2O3, α-FeOOH, and Fe3O4) with dispersed calcium ferrites (CaFe2O4 and Ca2Fe2O5) bonding the mineral grains [2,3]. As the reduction reaction progresses, cracks and holes are created in the sinters due to volume reduction of the iron oxide grains. Investigating the heterogeneous progress of reduction reaction leads to understanding how the quality of iron-ore sinters affect the performance of the blast furnace, and then realizing a lower-cost process of steel production.

Samples of three different reduction stages have been prepared for the experiment. 2×2×5 mm specimen cut from a bulk iron ores are first investigated with laboratory-source-based X-ray CT to specify the region of interest. The cracks and holes are found to increase as the reduction reaction progresses. The samples are then investigated with semi-μXAFS instrument. X-ray absorption of Fe K-edge and distribution of Ca have been mapped simultaneously by XRF detection.

Figure 3 shows the mapping of reduction stage of Fe for each reduction stages. Whiteout regions indicate the voids, that are holes and cracks. Heterogeneous reduction of iron from Fe(III) to Fe(II) was clearly observed. Round holes in Fig. 3(a) are found to exist from the initial stage of reduction, therefore not to be the reaction surface. In contrast, spiky holes and cracks in Fig. 3(c) are the voids created due to volume reduction, where the reaction has been finished around them. Thus the microstructure of iron-ore sinters might play an important role in reduction reaction. Further analysis would reveal the relationship between the dispersion of calcium ferrites and the heterogeneous reaction process.

(a) (b) (c)

250um

Fe(III)

Fe(II)


31

Fig. 3. Reduction stage mapping of iron-ore sinters. Bulk reduction stage progress from (a) to (c).

4. Conclusions We have developed a semi-μXAFS instrument that is capable of simultaneous XAFS/XRF/XRD mapping. It reveals the chemical heterostructures of structural materials with the spatial resolution of ~20 μm. As an example, the reduction process mapping of iron-ore sinters was presented. With further application for in situ experiments, this method provides fruitful information about heterogeneous propagation of deterioration phenomena in structural materials.

Acknowledgement A part of this work was supported by Cross-ministerial Strategic Innovation Promotion Program (SIP, unit D66) operated by the cabinet office. A part of this work has been performed under the approval of the Photon Factory Program Advisory Committee (Proposal No. 2015S2-002). A cooperative research between Nippon Steel & Sumitomo Metal Co. and KEK is also acknowledged.

References [1] N. Igarashi, N. Shimizu, A. Koyama, T. Mori, H. Ohta, Y. Niwa, H. Nitani, H. Abe, M. Nomura, T. Shioya, K.

Tsuchiya, and K. Ito, "New high-brilliance beamline BL-15A of the Photon Factory", J. Phys.: Conf. Ser., 425, 072016 (2013).

[2] S. N. Ahsan, T. Mukherjee, and J. A. Whiteman, "Structure of fluxed sinter", Ironmaking and Steelmaking, 10, 54-64 (1983).

[3] L. Lu, R. J. Holmes, and J. R. Manuel, "Effects of Alumina on Sintering Performance of Hematite Iron Ores", ISIJ International, 47, 349-358 (2007).


32

Keynote 2

Yutaka Kagawa


33

Concept of "Materials Integration" and Some Examples

Yutaka Kagawa 7-3-1, Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan

Abstract: Materials integration system is under development to predict the performance (life-time); integrate theories, experimental knowledge, computation, measurement, database etc.; to utilize big-data; and to establish R&D center, capacity building and global network.

Main subjects of Materials Integration System are i) to predict the performance (life-time) of elements/structure which are manufactured from various choices of materials and processes, ii) to integrate theories, experimental knowledge, computation, measurement, database etc., and to utilize big-data, iii) contributing to reduce development time, to realize efficient development, to reduce manufacturing cost, to optimize the selection of materials and processes, to improve the reliability prediction and to reduce diagnosis and maintenance cost, and also iv) aiming to establish R&D center, capacity building and global network [1].

Reference [1] http://www.jst.go.jp/pdf/pamph_sip_en.pdf.

Fig. 1. Concept of materials integration.


34

Invited talk

Mihail Ionescu


35

In-situ and ex-situ characterisation of radiation induced modifications in materials

M. Ionescu1), D. Bhattacharyya 1), D. Gregg 1), T. Wei 1), R. Aughterson 1), A. Reichardt 2), J. Davis 1), I. Karatchevtseva 1)

1) Australian Nuclear Science and Technology Organisation, New Illawarra Rd, Sydney, NSW 2234, Australia, phone +612 9717 3301, fax +612 9797 3257, [email protected]

Department of Nuclear Engineering, University of California Berkeley, Berkeley, 94720 USA

Abstract: Damage of nuclear materials under the influence of radiation can result in premature failure. For metallic structural components this can lead to an increase of strength and decrease of ductility, and for ceramic materials used for waste immobilisation or inert matrix for fuel, the self-irradiation can affect the structural and macroscopic properties. Some examples of radiation testing results will be presented. Sr0.5Zr2(PO4)3 and Ln2TiO5 with applications in nuclear waste sequestration were investigated for radiation stability by ex-situ and in-situ approach. A 45XD TiAl alloy with a lamellar microstructure was damaged using 5MeV He, between room temperature and 500oC, and the irradiation hardening was measured by nano-indentation. Small He accumulations were observed by TEM in both α2 and γ phases, which increased in size with irradiation temperature up to 300oC and then decrease with further temperature increase, whilst hardening was decreasing between room temperature and 500oC. The effect of ion irradiation on the tensile properties of pure Ni single crystals was investigated using an in situ micro-mechanical testing device inside a scanning electron microscope. A 12.8 μm-thick Ni single crystal was irradiated with 6MeV He ions to a peak damage of 19 displacements per atom (dpa). Micro-tensile samples were fabricated from the specimens parallel to <100> using a focused ion beam instrument, and tested in tension up to fracture. The peak strength increased from ~230MPa for the un-irradiated material to about 370MPa and 500MPa for the 10 dpa and 19 dpa samples respectively, while the ductility decreased with increasing dose. The surface near the peak damage layers fractured in a brittle manner, while the layers with smaller dose underwent significant plastic deformation. Slip bands extended to the peak-damage layer in the sample with a dose of 19 dpa, but did not propagate further. TEM confirmed the stopping of the slip bands at the peak-damage layer.


36

Analysis of hydrogen in materials with the 15N nuclear reaction

combined with thermal desorption spectroscopy

K. Fukutani Institute of Industrial Science, University of Tokyo

Abstract: Nuclear reaction analysis (NRA) via the 1H(15N,αγ)12C reaction using a high-energy 15N ion beam allows for quantitative and non-destructive depth profiling of hydrogen in materials. In combination with surface analysis techniques, behaviour of hydrogen in metals is investigated with NRA.

1. Introduction

Many materials have solubility for hydrogen, and such dissolved-hydrogen often affects the electric and mechanical properties of the host materials. Typical examples are hydrogen storage, optically switchable mirrors, hydrogen purification with membranes, H-induced conductivity and the embrittlement of technical alloys. Elementary processes of hydrogen dissolution into materials include dissociative adsorption, surface diffusion, surface penetration to subsurface sites, and bulk diffusion, which are normally temperature-dependent. Despite the importance of such hydrogen dynamical behaviour, it is rather elusive with standard analysis techniques, because hydrogen has only a single electron and interaction with electron or X-ray probes is weak. High-energy ion beam methods such as nuclear reaction analysis (NRA) and elastic recoil detection (ERD), on the other hand, allow for high-sensitive detection of hydrogen.

In view of growing interest in hydrogen-materials interaction, we have studied hydrogen interaction with surfaces of metals and metal oxides by using NRA via the resonant 1H(15N,αγ)12C reaction combined with conventional surface science techniques [1]. This resonant NRA determines hydrogen surface coverages and absorbed concentrations quantitatively, non-destructively, and with high depth resolution. In this paper, we briefly introduce this NRA technique and some applications to Pd [2-4], PdAu alloys [5], stainless steel [6], and metallic glassy alloys [7].

2. Experimental method For the detection of hydrogen with NRA via the 1H(15N,αγ)12C reaction, 15N ions are accelerated to the resonance energy of 6.385 MeV (ER), and γ rays at 4.43 MeV emitted as a result of the reaction are detected by scintillators. Since the resonance width of this reaction is as narrow as 1.8 keV, scanning the energy of the incident ion allows for depth profiling, which is schematically shown in Fig. 1. Whilst 15N ions at ER react with H at the surface of samples, ions at an energy of higher than ER react with H inside materials. The depth resolution near the surface is a few nanometers in normal incidence.

The experiments are performed with the tandem accelerator at MALT (Micro Analysis Laboratory) of University of Tokyo. The typical beam current is 50 nA at the sample. The ion beam is focused to 2 mm and <100 μm in diameter at the beam lines of 1E and 1C, respectively. At the 1C beam line, NRA measurements under the atmospheric condition are possible by extracting the beam through a SiN membrane [8].

Samples are prepared either ex situ or in situ in an ultra-high vacuum (UHV). When hydrogen at the top-most surface is concerned, the sample surface is cleaned in UHV, and analyzed with NRA in situ. In order to examine the thermal stability of hydrogen at a specific depth, furthermore, thermal desorption spectroscopy (TDS) is also applied in combination with NRA.

3. Results Pd is well-known as a hydrogen-absorbing material. While the diffusion kinetics of hydrogen in bulk has been studied in detail, the initial step for hydrogen dissolution, surface to subsurface penetration, is yet to be elucidated. Figure 1 shows the NRA yield curve representing the depth profile of H taken after exposure of a clean Pd(100) surface to H2 in UHV [2]. The curve reveals a maximum at ER corresponding to the hydrogen adsorbed on the surface. From the integrated intensity of this component, the hydrogen coverage is estimated to be 1 monolayer. In addition to this surface H, another component is present at an energy of higher than ER, which indicates some H is present below the surface. Figure 2 shows the TDS spectrum and the temperature dependence of the NRA signal intensity at specific

ER

ER+Δ E

γ yi

eld

ER ER+Δ E15N Energy (E0)

Fig. 1 Schematic of H depth profiling.


37

depths [2]. TDS reveals two maxima at 180 and 330 K. When NRA probes H at a depth of 6 nm, it was found that the intensity drops at a temperature of 180 K as shown by the blue curve in the figure. When H at the surface is detected, on the other hand, the signal intensity was reduced at about 330 K. These results unambiguously indicate that the TDS peaks at 330 and 180 K correspond to hydrogen at the surface and below the surface, respectively.

Figures 3 shows the amount of H absorbed at Pd(110) as estimated by TDS as a function of hydrogen-exposure temperature [3]. With increasing substrate temperature, H was more efficiently absorbed into Pd(110). By assuming the Arrhenius relation, the activation barrier for penetration was estimated to be 0.03 and 0.06 eV for H and D, respectively. These values are an order of magnitude smaller than the adsorption energy of hydrogen at the surface. On the basis of these results, it was proposed that surface penetration of hydrogen is a concerted process involving another hydrogen molecule.

Stainless steel is a material widely used in many applications. Hydrogen dissolution and its thermal stability are of particular interest in terms of vacuum technology and mechanical properties. A type 304 stainless steel, of which composition was Fe:71.6, Cr:18.6, Ni:8.4, Mn:1.0, Cu:0.1, V:0.1, and Mo:0.1 in mol% as confirmed by X-ray fluorescence analysis, was analysed by NRA at various temperatures. Figure 5(a) shows the NRA yield curves measure at elevated temperatures from 300 to 750 K (the data indicated as 975 K was measured at room temperature after heating at 975 K). The curve at 300 K reveals a maximum at an energy higher than ER by 3 keV. This peak feature is as broad as 17 keV accompanied by a tailing extending to a depth of ~10 nm. This indicates that the near-surface region with a thickness of ~10 nm contains a substantial amount of hydrogen. This probably corresponds to hydrogen in a metal oxide layer. In a deeper region, furthermore, there exists a significant amount of hydrogen, which seems to correspond to bulk-dissolved H. Fig. 5(b) shows the average γ-ray yield in the bulk region as a function of temperature. Apparently, the volume density of hydrogen in a deeper region decreases upon heating and remains almost constant above 550 K. The result implies that there are at least two types of hydrogen in this depth region: One is removed by heating up to 550 K and the other remains above 550 K. Fig. 5(c) shows the integral of the γ-ray yield for the surface component as a function of temperature. The area density of the surface hydrogen is 1.3x1016 cm-2 at 300 K. Then it gradually decreases above 400 K and remains constant even at 975 K. From the analysis of the data, we derived the distribution of the binding energy of H in the sample. It was found that the binding energy is distributed above 1 eV.

4. Conclusions The technique of NRA via the 1H(15N,αγ)12C reaction was briefly described. By applying this NRA method combined with thermal desorption spectroscopy, hydrogen penetration into Pd was investigated. Temperature dependence of the NRA was measured for a stainless steel sample. The binding of H in stainless steel was discussed.

Fig. 3 TDS and NRA yields at surface and 6 nm below the surface as a function of temperature [3].

Fig. 4 Amount of H absorbed in Pd(110) at various temperatures [3].

Fig. 2 NRA yield curve for Pd(100) [2].

Fig. 5 NRA measured for a stainless steel [6].


38

Acknowledgement

The studies presented here were accomplished in collaboration with M. Wilde, S. Ohno, K. Takeyasu, S. Ogura, D. Sekiba, and H. Matsuzaki. This work was supported by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS).

References [1] M. Wilde, K. Fukutani, “Hydrogen detection near surfaces and shallow interfaces with resonant nuclear reaction

analysis”, Surf. Sci. Rep. 69 (2014) 196.

[2] M. Wilde, K. Fukutani, “Penetration mechanism of surface adsorbed hydrogen atoms into bulk metals: Experiment and model”, Phys. Rev. B 78 (2008) 115411.

[3] S. Ohno, M. Wilde, K. Fukutani, “Novel insight into the hydrogen absorption mechanism at the Pd(110) surface”, J. Chem. Phys. 140 (2014) 134705.

[4] S. Ohno, M. Wilde, K. Fukutani, “Desorption temperature control of palladium-dissolved hydrogen through surface structural manipulation”, J. Phys. Chem. C 119 (2015) 11732.

[5] S. Ogura, M. Okada, K. Fukutani, “Near-surface accumulation of hydrogen and CO blocking effects on a Pd-Au alloy”, J. Phys. Chem. C 117 (2013) 9366.

[6] K. Takeyasu, M. Matsumoto, K. Fukutani, “Temperature dependence of hydrogen depth distribution in the near-surface region of stainless steel”, Vacuum 109 (2014) 230.

[7] D. Sekiba et al., “Development of site-specific NRA for hydrogen mapping: observation of fatigue-fractured surface of glassy alloy”, Nucl. Instr. Meth. B 269 (2011) 627.

[8] H. Yonemura et al., “Hydrogen depth profiling in an atmospheric pressure”, Nucl. Instr. Meth. B 269 (2011) 632.


39

Life Cycle Monitoring and Quality Assessment of Advanced Polymer Matrix Composites

Nobuo Takeda1), Shu Minakuchi 2) 1) TJCC (UTokyo-JAXA Center for Composites), Graduate School of Frontier Sciences, The University of Tokyo,

Mail Box 302, 5-1-5 Kashiwanoha, Kashiwa-shi, Chiba 277-8561, Japan, [email protected] 2) [email protected]

Abstract: Optical fiber sensors are very useful to monitor the internal strain and temperature in composites during manufacturing as well as in practical operations. The authors have been using both multi-point and distributed strain monitoring techniques to characterize the internal state of composite structures. This paper reports some recent developments of our work. Specifically, distributed sensing for large-scaled parts, through-thickness strain monitoring for complex-shaped parts, and direction-dependent cure shrinkage monitoring are described, highlighting wide applicability of embedded optical fiber sensors for intelligent process/life cycle monitoring and quality assessment of advanced polymer matrix composite parts.

1. Introduction Even though carbon fiber reinforced plastic (CFRP) is being used in almost all modern aerospace structures as a primary structural material, it is still difficult to precisely manufacture large-scale CFRP structures and ensure their structural integrity during operation. Hence, there is an urgent need to develop innovative sensing techniques to monitor the internal states of composite structures and utilize the obtained data to improve safety, structural design, processing technologies and maintenance methods. Within the systems developed so far, optical fiber sensors have attracted considerable attention [1], since they are small, lightweight, immune to electromagnetic interference, and environmentally stable. Furthermore, they possess sufficient flexibility, strength, and heat resistance to be embedded relatively easily into composite laminates during the fabrication. Embedded sensors continuously monitor the composite manufacturing process itself, in-service usage, and damage (Fig. 1, [2]). By combining all the information obtained by the sensing network, we can accurately evaluate the internal state of the structure. This paper reports some recent developments of our work, highlighting wide applicability of embedded optical fiber sensors for intelligent process monitoring and quality assessment of composite parts.

2. Distributed sensing for large-scale structures Reference [2] demonstrated fiber-optic-based life cycle monitoring of a representative CFRP stiffened panel (Fig. 2) manufactured by vacuum assisted resin transfer molding (VARTM). A single optical fiber was embedded between the stiffeners and the skin during the laminate lay-up process and the formed fiber-optic network was then utilized to monitor the manufacturing process and subsequent impact tests. A Brillouin-based system with a spatial resolution of 10 cm was utilized for distributed strain measurement. The internal state of the panel was successfully monitored throughout its life (Fig. 3), confirming the effectiveness of life cycle monitoring by fiber-optic-based distributed sensing for developing highly-reliable composite structures.

Fig. 1 Life cycle monitoring [2].


40

Fig. 2 CFRP panel for demonstration [2].

Fig. 3 Thermal residual strain obtained from two lines embedded in the same position. Result agreed well with FBG measurement, validating measurement accuracy [2].

Following this demonstration, a hybrid Brillouin-Rayleigh system was developed, which can measure strain and temperature distribution simultaneously [3]. The system could identify a non-uniform thermal residual strain field induced by a non-uniform cure temperature. The distributed monitoring approach was further utilized to clarify the mechanism of strain development in a thick composite panel curing under a non-uniform temperature in the thickness direction [4]. It was confirmed that viscoelasticity is the key of the strain development.

3. Through-thickness strain monitoring using birefringent effect of fiber Bragg grating A fiber Bragg grating (FBG) sensor has a periodic variation in the refractive index along the length of a single mode optical fiber. When broadband light is launched into the FBG sensor, a narrow spectral component is reflected back, and the reflection spectrum gives the measure of strain and/or temperature. When a non-axisymmetric strain state arises at the core of the FBG sensor (i.e., cross-sectional shape of the sensor is flattened), the reflection spectrum from the sensor splits into two peaks due to the birefringence effect. The authors have utilized this spectral response to measure through-thickness strain in complex-shaped composite parts.

One example is an L-shaped composite part [5]. L-shaped parts are structural key elements in complex-shaped aerospace composite structures, and life cycle cross-sectional strain change is their key. An FBG sensor was embedded in the through-thickness center of a corner part and the cross-sectional strain change was continuously evaluated by using a birefringence effect of the FBG. The embedded sensor could capture key strain changes throughout the structural life cycle with assistance of a new optical fiber reconnection method (Fig. 4). It was confirmed that ‘‘non-axisymmetric strain’’ of the FBG sensor is a good indicator for spring-in distortion in manufacturing and through-thickness tensile failure in operation. The developed system was further utilized for advanced quality assurance of assembled L-shaped parts, and a new strength prediction method was proposed based on the FBG response.

This technique was then applied to CFRP pipes [6]. In thick-walled composite pipes, significant radial tensile stress is induced during curing, leading to premature delamination failure and performance degradation. The fiber-optic-based system could sensitively capture through-thickness stress development and detect delamination failure in pipes. Following the establishment of the monitoring technique, a residual stress reduction method was proposed to develop thick-walled crack-free CFRP pipes [7]. The authors began by addressing the effect of stacking sequence on residual stress using theoretical and numerical analyses. The result led us to a novel stress-reduction method where circumferentially stiff layers are gathered close to the inner surface. Two pipes were then manufactured: asymmetric and symmetric lay-ups. The radial strain development was evaluated using the fiber-optic-based monitoring system to


41

Fig. 4 (Left) Life cycle monitoring of L-shaped corner part. Key change in through-thickness strain was continuously monitored using birefringence effect of FBG. (Right) New reconnection technique critically important for manufacturing and reparing composite parts with embedded optical fiber [5].

Fig. 5 Cross-section of thick-walled CFRP pipes. No delamination failure was observed in proposed asymmetric lay-up. Stress reduction effect was validated using fiber-optic-based monitoring system [7].

Fig. 6 (Left) Smart boltless structure with embedded optical fiber sensor. Without bolts, component thickness can be reduced and more efficient load path is possible. (Right) FBG spectrum just after vacuuming. Using this response, manufacturing workers can immediately identify areas with lack of bonding pressure [8].

confirm the effectiveness of the method. Finally, a thick-walled pipe was fabricated. No failure was observed during curing, successfully demonstrating a thick-walled, crack-free CFRP pipe (Fig. 5).

The FBG-based approach was further extended to process control of composite bonded joints [8]. Stringent process control of bonding is one of the keys to realization of bonded joints in composite aircraft primary structures (i.e., realization of boltless structures) (Fig. 6). The authors developed the first fiber-optic-based quality control technique. Bonding pressure and residual strain in the adhesive were monitored, which enables manufacturing workers to identify areas with lack of bonding pressure before curing and to assure the quality of the bonded joint after curing. The proposed technique will be a key technology to improve the reliability of bonded joints and to reduce the manufacturing cost.


42

Fig. 7 (Left) Surface of specimen with through-thickness sensor. Optical fiber is extending perpendicularly from surface. (Right) Transversely anisotropic shrinkage of unidirectional carbon/epoxy induced due to tool-part interaction. Measurement under practical curing condition is possible with this approach [9].

4. Embedding sensors in through-thickness direction Reference [9] developed a fiber-optic-based technique for in situ characterization of direction-dependent cure-induced shrinkage in thermoset fiber-reinforced composites. A procedure was established to embed fiber Bragg grating (FBG) sensors in composite out-of-plane directions and to measure key through-thickness chemical cure shrinkage directly under practical curing conditions (Fig. 7). FBG sensors embedded in through-thickness and in-plane directions clarified direction-dependent cure-induced shrinkage in autoclaved unidirectional carbon/epoxy. The developed technique will be a powerful tool for evaluating cure shrinkage in complex-shaped parts and for validating process-simulation tools based on internal strain.

5. Conclusions This paper reported some recent developments of our work. Specifically, distributed sensing for large-scaled parts, through-thickness strain monitoring for complex-shaped parts, and direction-dependent cure shrinkage monitoring were described, highlighting wide applicability of embedded optical fiber sensors for intelligent process/life cycle monitoring and quality assessment of composite parts.

Acknowledgements This paper is supported by both JSPS KAKENHI Grant Number 26220912 and “Innovative Structural Material Project”, Cross-ministerial Strategic innovation Promotion Program (SIP), Japan.

References [1] S. Minakuchi, and N. Takeda “Recent advancement in optical fiber sensing for aerospace composite structures,”

Photonic Sensors, 3 (4), 345-354 (2013). [2] S. Minakuchi, N. Takeda, S. Takeda, Y. Nagao, A. Franceschettic, and X. Liu, “Life cycle monitoring of large-

scale CFRP VARTM structure by fiber-optic-based distributed sensing,” Composites Part A: Applied Science and Manufacturing, 42(6), 669-676 (2011) .

[3] Y. Ito, S. Minakuchi, T. Mizutani, and N. Takeda “Cure monitoring of carbon-epoxy composites by optical-fiber-based distributed strain-temperature sensing system,” Advanced Composite Materials, 21(3), 259–271 (2012).

[4] Y. Ito, T. Obo, S. Minakuchi, N. Takeda, “Cure strain in thick CFRP laminate: optical-fiber-based distributed measurement and numerical simulation,” Advanced Composite Materials, 24(4), 325-342 (2015).

[5] S. Minakuchi, T. Umehara, K. Takagaki, Y. Ito, N. Takeda “Life cycle monitoring and advanced quality assurance of L-shaped composite corner part using embedded fiber-optic sensor,” Composites Part A: Applied Science and Manufacturing, 48: 153–161 (2013).

[6] K. Takagaki, S. Minakuchi, N. Takeda, “Fiber-optic-based life cycle monitoring of through-thickness strain in thick CFRP pipes,” Advanced Composite Materials, 23(3), 195–209 (2014).

[7] K. Takagaki, S. Minakuchi, N. Takeda, “Thick-walled crack-free CFRP pipes: stress reduction using atypical lay-up,” Composite Structures, 126, 337–346 (2015).

[8] S. Minakuchi, K. Uhira, Y. Terada, N. Takeda, N. Saito, T. Shimizu, "Quality control of composite bonded joints using fiber-optic-based process monitoring," SAMPE Journal, 51(1), 44-51 (2015).

[9] S. Minakuchi, “In-situ characterization of direction-dependent cure-induced shrinkage in thermoset composite laminates with fiber-optic sensors embedded in through-thickness and in-plane directions,” Journal of Composite Materials, 49(9), 1021-1034 (2015).


43

Overview of theme 2: Quantitative elemental analyses of hydrogen and light elements in structural

materials

Eiji Kita 1), 2), Kimikazu Sasa 1), 2), Daiichiro Sekiba 1), 2), Tetsuaki Moriguchi 1), 2), Akiyoshi Yamazaki 2), Shigeo Tomita2), Masanori Kurosawa3), Satoshi Ishii1), Hiroshi Naramoto1), Masahiro Ukibe4), Sigetomo Shiki4), Takeshi Fujii4), Paul Fons4), Hideaki Kitazawa5),

Kazuhiro Hono5) 1) Tandem Accelerator Complex, Research Facility Center for Science and Technology, University of Tsukuba, 1-1-1 Tennodai,

Tsukuba, Ibaraki, 305-8577, Japan 2) Graduate School of Pure and Applied Sciences, University of Tsukuba,

3) Graduate School of Life and Environmental Sciences, University of Tsukuba 4) Res. Institute for Measurement and Analytical Instrumentation, National Institute of Advanced Industrial Science and Technology,

Tsukuba, Ibaraki 305-8568, Japan 5) National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan

Abstract: In performances and properties of structural materials, light and/or trace elements have played important roles, for example hydrogen brittleness in steels. The light elements are “ubiquitous” in circumstance and might be incidentally included in materials. They sometime cause an unexpected degradation in the performance. Quantitative evaluations of the amount and distribution of light and trace elements in structural elements will avoid such degradation. For this purpose, we are designing accelerator materials analysis techniques and are improving characteristic X-ray detection techniques. The precise elemental analysis including hydrogen under working conditions will give us keys for solving current problems as well as basic data like diffusion constants used in materials integration. 1. Introduction In the performances and properties of structural materials, the roles of light elements and small amount of addition elements have been playing important roles. Light elements exist everywhere in the environment and it is easy to consider that the materials include eventually them, however the amount of these light elements are not easy to analyse with enough quantitatively due to the difficulty in the detection techniques. Minor constituents in the materials have important roles in the performance. It can be said that the devilment of high performance alloys can be obtained by choosing proper minor additional elements. On the other hand, small amounts of unexpected additive elements sometime lead the performance to worth direction. As the development of analytical techniques, the detection of additional elements has been improved and this developments helps to control the performance of the materials. In this SIP project, we are working with structural materials which are used for aircrafts and turbine blades for electricity power generators. The major developing materials expected in the latest aircraft are the subject of the SIP project and categorized in the three fields, carbon fiber-reinforced composite plastics material(CFRP) to the airframe structural member, a composite material for a light-weight ceramics coating for jet engine turbine blades and a Ti alloy. To investigate the performance, fundamental issues to determine their life time, it is essentially important to survey at their working conditions. Mission of the theme 2 in the advanced structural materials SIP project is the quantitative elemental analysis. The elements lighter than oxygen are considered to be elements hard to analyse because the energies of their characteristic X-ray are lower than 1 keV. Popular analytical methods rely on the detection of characteristic X-ray and common difficulties in detection of low energy X-ray make the light element analysis hard. In this theme, the challenge to extend the range of detectable elements will performed as well as the developments of specific analytical techniques for light elements. Elemental analysis by using an ion accelerator used in a nanotechnology field is arranged to use for structural materials. Additionally high sensitive elemental analysis and 2 dimensional elemental mapping will be items to be developed.

2. Development of light elemental analitical techniques One of our special techniques is the elemental analysis using the ion accelerator, which has been used nanotechnology field such as materials analysis of electronic devices. There are several techniques and some of them are based on the measurement principle completely different from the way detecting characteristic X-ray. In these techniques, we used the interactions between high energy ion beams against materials. In Fig. 1, the interaction between high energy ion beams and materials is schematically illustrated. The ion beams interact with atoms in the material and cause the change in the state; collision and recoil between elastic spheres, nuclear reactions and excitation of electronic states. The first one is known as Rutherford back scattering(RBS) which is a popular elemental analysis technique for semiconductor and other nano-materials. This technique can measure hydrogen and is extremely sensitive to the surface atomic structure and this characteristic may not fit to the structural analysis. Elastic Recoil Coincidence Spectrometry


44

(ERCS) detects recoiled hydrogen atoms by 10 MeV proton beams. Nuclear Reaction Analysis(NRA) is based on the nuclear reaction between 15N having energies higher than 6.5 MeV and hydrogen atoms according to the formula, 15N + 1H γ-ray (4.9MeV) + 12C +α-Ray. The energy of 15N for the reaction is limited to one more than 6 MeV and the depth profile can be obtained by changing incident energy of accelerated ions. Fig. 1. Schematic illustration of usage of an ion accelerator and interaction between high energy ion beam and material. Each interaction can be used for an advanced elemental analysis for structural materials. The principle of accelerator mass spectroscopy is also schematically illustrated. It is noted that the accelerator mass spectroscopy (AMS) has the highest sensitivity against additional trace elements. Particle Induced X-ray Emission spectroscopy (PIXE) is also operated by using ion accelerators. The characteristic X-ray is emitted by high energy particle instead of electrons in Energy Dispersive X-ray spectrometry (EDX). This technique has advantages where the lower back ground compared with EDX and simultaneous multi-elemental analysis. We are designing to attach the high sensitive superconducting X-ray detector[1] developed in AIST, which works for the light element analysis, for example boron. We are also constructing a micro ion beamline for PIXE. Currently, the ion accelerator to be used in this project is building at the Tandem Accelerator Complex, University of Tsukuba. The 6 MV tandem-type electrostatic accelerator will be available at the end of year. The efforts to improve the sensitivity for lower energy characteristic X-ray and ordinal mass spectroscopy are concentrated for transmission electron micrograph (TEM) analyses and time-of flight secondary ion mass spectroscopy(TOF-SIMS) in National Institute for Materials Science. These are effective to extend the present analysis techniques to light element analysis.

3. Summary and future plan For the quantitative evaluation of the amount and distribution of light and trace elements in structural elements, we are designing accelerator materials analysis technique and are improving characteristic X-ray detection. The precise elemental analysis including hydrogen under working conditions will give us keys for solving current problems as well as basic data like diffusion constants used in materials integration.

Acknowledgements This work is supported by the Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.

References [1] S. Shiki, M. Ukibe, Y. Kitajima, M. Ohkubo, “X-ray detector performance of 100-pixel superconducting tunnel

junction array detector in the soft X-ray region”, J. Low Temp. Phys., 167, 748-75


45

IBA for Hydrogen Uptake Observation in Functional Metals under Ambient Condition

Daiichiro Sekiba University of Tsukuba


46

An XAFS Structural and Chemical Analysis Approach for Trace Light Elements of B and N in Heat-Resistant Steel

P. Fons, S. Shiki, M. Ohkubo National Institute of Advanced Industrial Science & Technology (AIST), 1-1-1 Higashi, Tsukuba, 305-8573, Japan

+81-29-861-5636, [email protected]

Abstract: Fluorescence-based measurements of the local structure about N atoms in high-Chromium heat-resistant ferric steels (wt.% 0.08 C, 0.3 Si, 0.5 Mn, 9 Cr, 0.2 V, 0.06 Nb, 0.015 B) were carried out for two N concentrations using a 100-pixel superconducting-tunnel-junction array detector. X-ray Absorption Near Edge Structure (XANES) spectra of the N impurity in steel were successfully acquired for the first time. For a 0.007 N wt% alloy , N was predominated incorporated interstitially into the bcc ferrite matrix. For the larger N concentration (0.03 wt.%) alloy, N was incorporated into a mixture of interstitial ferrite sites and the cubic phase of VN. No evidence for N incorporating in BN was found. While B fluorescence was detected in the samples, further improvements in detector sensitivity are required to acquire XANES spectra of the B impurity at meaningful statistical quality levels.

1. Introduction High Cr ferritic heat resistant steels, such as martensitic 9 to 12% Cr steels, have long been used for steam pipe applications in conventional thermal power plants with maximum steam temperatures of 620 oC. To improve overall efficiency and at the same time reduce the amount of the greenhouse forming gas CO2 generated from thermal power plants during operation, materials-development projects for advanced ultra-supercritical (A-USC) power plants with steam temperatures of 700 °C and above have been carried out worldwide. These projects all involve the replacement of martensitic 9 to 12% Cr steels with Ni-base alloys for the highest temperature boiler and turbine components in order to ensure sufficient creep strength. To minimize the requirement for expensive Ni-base alloys, martensitic 9 to 12% Cr steels can be applied to the next highest temperature components with temperatures of 650 oC or below in A-USC power plants.

Critical issues for the development of advanced martensitic 9 to 12% Cr steels for 650 oC boilers are the stabilization of microstructure at and near prior austenite grain boundaries (PAGBs) for long times at 650 oC [1]. Preferential recovery of martensitic microstructure in the vicinity of PAGBs promotes localized creep deformation in the vicinity of PAGBs, which results in premature creep rupture [2]. A fine distributions of precipitates such as M23C6 carbides and MX carbonitrides at and near PAGBs enhances grain boundary precipitation hardening and suppresses preferential recovery of martensitic microstructure in the vicinity of PAGBs. The addition of small amount of boron at around 100 ppm stabilizes fine distributions of M23C6 carbides at and near PAGBs for long times at 650 oC, which results in a significant increase in long-term creep rupture strength at 650 oC. It has also been proven that nitrogen is beneficial to the long-term creep strength of ferritic and austenitic steels through solid solution hardening as well as precipitation hardening by the formation of fine nitrides, such as fine vanadium nitrides (VN). The addition of boron and nitrogen without the formation of any BN during normalizing heat treatment significantly improves the creep strength but excess addition of boron and nitrogen result in the formation of BN during normalizing heat treatment, which consumes soluble boron and nitrogen and hence degrades the creep strength [3]. Here we investigate the local environment about nitrogen atoms for two 9% Cr steels containing high boron of 130 to 150 ppm but different nitrogen concentrations using X-ray absorption near edge spectroscopy (XANES).

2. Experiment and Purpose 2.1 Experimental Procedure Two 9% Cr steel specimens with the chemical compositions shown in Table 1 were used for the XANES experiments. The specimens, having a size of 10 mm diameter and 2.5 mm thickness, were taken from heat treated 9% Cr steel. The specimens were polished and then placed into a sample holder at beamline bl-16a at the Photon Factory in Tsukuba. The beamline was equipped with a APPLE-II type undulator and a varied-line-spacing monochromator. For the current experiment, the polarization of the incident x-ray beam was in the plane of the samples. The estimated photon flux was about 3 x 10 11 photons/second at the Nitrogen K-edge, although the incident flux was estimated to be a factor of ten lower for the B K-edge. The samples were irradiated with a ten degree incident beam to enhance the fluorescence intensity. B (188 eV) and N ( 410 eV) K-edge XANES spectra were taken from each sample after appropriate energy windows were defined on the 100-Pixel superconducting-tunnel-junction array detector (STJA) to detect the corresponding fluorescence from B (K-L2 183.3 eV) and N (K-α2 392.4 eV) fluorescence lines, respectively. Data were integrated to yield better than 1% standard error assuming Poisson statistics. 2.2 Superconducting-tunnel-junction array detector


47

The soft x-ray energy range below 1 keV is of importance for detecting the local structure about light elements. Due to the close proximity of the K-edges as well as M and higher order lines from heavier elements, high energy resolution is a crucial parameter for a fluorescence detector in this range due to the possibilities of interference from other nearby lines originating from elements not under study. A typical semiconductor-based detector such as a Si drift detector has an energy resolution of more than 100 eV (FWHM) making it inappropriate to distinguish a 0.007 wt% B fluorescence signal. In contrast, the energy resolution of the STJA detector is approximately 11 eV and hence has sufficient resolution to completely reject fluorescence from neighboring atoms in the periodic table even for the case when the concentrations of background atoms is larger than the element under study [4].

For the current experiment, while it was demonstrated that B K-edge fluorescence can be detected for higher concentration samples, the combination of an order of magnitude drop in incident intensity of the beamline in going from the N XANES energy region to that of B in combination with the reduced fluorescence efficiency of the lighter element B in comparison to N, made it not feasible to collect B XANES from the two heat-treated steel samples with the current revision of the STJA detector.

Table 1. Chemical compositions of 9% Cr steels examined.

3. Data and Results Figure 1 shows the N K-edge XANES signals for the two samples S1 and S2. A clear trend in the shape of the N K-edge XANES spectra can be seen with increasing N from S2 (N 0.007 wt%) to S1 (N 0.030 wt%). This difference becomes clearer in the difference spectra shown in Figure 2.

Figure 1. XANES N K-edge spectra for samples S1 and S2 taken with the superconducting-tunnel-junction array detector.


48

In addition to the S1-S2 difference spectra, two reference spectra are shown in Figure 2, the corresponding N K-edge spectra for VN (from ref. [5]) and for BN (measured along with samples S1 and S2. Although an experimental reference, is not available, N at low concentrations is thought to incorporate into the body-centered cubic ferrite phase. Theoretical XANES spectra derived from ab-initio density functional theory simulations of tetrahedral and octahedral interstitial N are shown at the bottom of Figure 2. The position of N atoms in isolated tetrahedral and octahedral interstices in the ferrite phase was determined using the ab-initio plane wave code CASTEP[6]. The relaxed positions were in turn used as input to the FDMNES code from which theoretical XANES spectra were calculated by solving the Schrodinger equation using finite difference techniques [7].

Figure 2. Difference curve (S1-S2) of N K-edge signals for samples S1 and S2 (black) along with h-BN and VN N K-edge XANES reference samples, and a theoretical XANES spectra for N incorporated into a weighted sum of tetrahedral and octahedral interstices. 5. Discussion Semba and Abe reported that most of nitrogen in 9Cr steel with low nitrogen of 0.0079% is in solid solution after tempering heat treatment, while that in 9Cr steel with high nitrogen of 0.03 and 0.06% precipitates as nitrides such as BN and VN after tempering [8]. Referring to the difference curve, it is clear that the experimental signal can be best approximated by a sum of the two reference XANES spectra VN and interstitial N in ferrite. This suggests that for sample S2, the N is largely present as interstitially incorporated N in the ferrite matrix. The high nitrogen of 0.03% in the specimen S1leads to the formation of VN and BN in the matrix. No evidence for the formation of h-BN is found.

In addition to the discussion of N incorporation in heat-resistant steels, we shall also demonstrate the successful measurement of B fluorescence from the steel samples with pulse height distribution data from the STJA detector. While the intensity of the B fluorescence data was insufficient to acquire B K-edge XANES spectra for the current experiment, we shall discuss possible changes to the STJA detector as well as proposed changes to the incident optics that may make such spectra possible in the near future.

7. Conclusions N K-edge XANES were, for the first time, acquired in fluorescence mode from two heat-resistant 9 Cr steel samples with N concentrations as low as 0.007 wt%. The N K-edge difference spectra between the higher N concentration


49

sample S1 and the lower concentration sample S2, indicates that the changes can be largely attributed to the formation of a new cubic VN phase and a background signal ascribable to the presence of interstitial N in the ferrite phase matrix. These conclusions are consistent with the proposed model by Yamada et al. [7].

While B K-edge fluorescence mode XANES was successfully acquired from a h-BN reference sample, the B fluorescence signal from samples S1 and S2 was too weak to obtain viable XANES spectra, although a pulse height distribution analysis of the STJA detector output for a 193 eV incident beam showed a clear B signal. In the future to be able to measure B K-edge XANES from samples with B concentrations comparable to the current samples (0.01 wt%) will require at least a factor of 100 improvement in detector sensitivity and or sampling. As the thin polymer shielding present in the current detector design result in large absorption of the B fluorescence signal, replacing the shielding with alternative materials is expected to lead to a factor of ten improvement in sensitivity. This, in conjunction with optimization of the incidence geometry may provide sufficient intensity to acquire B K-edge XANES fluorescence data from samples such as those used in this experiment in the future.

Acknowledgements The authors would like to thank F. Abe for useful discussions. This work was supported by the Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.

References [1] F. Abe, Science and Technology of Advanced Materials, 9 (2008) 013002. [2] H. Kushima, K. Kumura, F. Abe,Tetsu-to-Hagane, 85, 841 (1999) (in Japanese). [3] F. Abe, M. Tabuchi and S. Tsukamoto, Energy Materials, 4 (2012) 166-175. [4] M. Ohkubo, S. Shiki, M. Ukibe, N. Matsubayashi, Y. Kitajima, and S. Nagamachi, Sci. Rep. 2, 831 (2012). [5] J.G. Chen, J. Eng Jr., and S.P. Kelty, Catalysis Today 43, 147 (1998) [6] Stewart J. Clark, Matthew D. Segall, Chris J. Pickard, Phil J. Hasnip, Matt I. J. Probert, Keith Refson and Mike C.

Payne, Z. Kristallogr. 220, 567 (2005) [7] Y. Joly, Phys. Rev. B 63, 125120 (2001). [8] H. Semba and F. Abe, Energy Materials, Vol.1, No.4 (2006) 238-244.


50

Nanoscale Characterization in Structural Materials by SAXS and SANS

Masato Ohnuma1), Michihiro Furusaka 1), B.S.Seon2), Toshinori Ishida1) , D.H.Ping 3) 1) Hokkaido University, K13-N18, Sapporo, Hokkaido 060-8628, Japan, +81-11-706-6650, [email protected]

2) Korean Atomic Energy Research Institute (KAERI), Daejeon, South Korea 3) National Institute for materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Japan

Abstract: Small-Angle Neutron Scattering (SANS) profiles measure by newly designed instrument installed in the in-house neutron source have been presented with comparing the data obtained from well established SANS instrument installed in research reactor. Both data are overlapped nicely, indicating usefulness of small scale neutron source in materials Science.

1. Introduction Small-Angle Scattering technique can bring us quantitative information about the average size, number density and volume fraction of precipitates. Because of simple transmission geometry, measurement itself is rather easy. No complicated sample preparation is required, just make the sample with proper thickness for getting enough transmission rate. For most of the steels, the typical sample thickness for SAXS and SANS is 20~50 μm and 1~2 mm, respectively. In addition, once the proper model for fitting is assigned, analysis is not at all difficult. Therefore, the technique can be a potentially candidate for “non-destructive”, “quick and easy” nano-characterization tool, especially in the case of SANS. The number of SANS instruments, however, is quite limited: i.e. only a few SANS in each facility and they are always over subscribed. Most of the SANS facility can measure of more than three-decade of q range (q=4πsinθ/λ: λ is wave length of the beam) which corresponding to the scale range from a few nano-meter to a few hundred nano-meter. This requires a large size of instruments that typically reach around 20 m length. One of the authors (M.F.) suggests a possibility of medium range SANS instrument which can be used with small scale neutron facility such as Hokkaido University Neutron Source (HUNS) where neutrons are generated by 45 MeV electron Linear Accelerator (LINAC). In this paper, we report the performance of the new medium range SANS by showing the analysis of precipitates with a few nano-meter in diameter formed in high nitrogen martensitic steels and compare the data to those obtained in conventional SANS.

2. Results and Discussion The studied alloy is high nitrogen martensitic steel with the composition of 0.1C-16Cr-1Mo-0.6N in wt.%. This steel shows a good corrosion resistance and high hardness (Hv~670) when the steel tempered at 450ºC-1h, while the corrosion resistance is degraded after the tempering over than 500ºC [1]. One of the authors (M.O.) previously reported that the mechanism of hardness increment by tempering is formation of nano-size precipitate (or clusters) using SANS and SAXS [2]. M.O. also reported that combination of SANS and SAXS can give phase and compositional information when both profiles obtained in absolute intensity scale [3]. However the q range of SANS in the previous study is not high enough because the size of precipitates formed in the sample tempered at 450 C is only 1 nm in a diameter. Corresponding q-range should be around 1 nm-1. In contrast, most of SANS instruments for dedicate to measure at least 0.01nm-1 and relatively light weight for high-q measurements. For connecting the gap between demand from nano-science & engineering and from the “common-sense” in SANS, our target is set up to measure q range from 0.1 to 7 nm-1. For this purpose, we specially designed for limited q-range with using pulsed neutron source in HUNS by optimizing the q-resolution from typical value (2~3 mrad) to relaxed value (~10 mrad) which is enough high resolution for our target. Since our nano-SANS is specialized for nanostructure analysis in steel, we call our “intermediate angle neutron scattering” instrument as “iANS (irons)”. Meanwhile, we also measured exactly same samples using 18m SANS installed in the cold source of HANARO in Korea Atomic Energy Institute (KAERI). The measurements have performed and analysed independently and compare profiles in absolute unit intensity. The measurements in HUNS have performed in the magnetic field in 0.6 T using Nd-Fe-B permanent magnets and in 18m SANS, the magnetic field is 1.0 T using electro magnet. For scaling to absolute unit, the glassy carbon has been used for both SAXS and SANS in HUNS and SiO2 in HANARO. 2.1 Performance of iANS evaluated by comparing to18m SANS By using the scattering angle 2θ from 5 to 15o and the wavelength from 4 to 10 Angstrom, the q range from 0.2 to 7 nm-

1 is covered by iANS. This is almost same q-range that can covered by our labo-SAXS with Mo-Kα source (0.1 to 10 nm-1). Therefore, we can cover enough q-range by our own facilities of SANS and SAXS, which is very advantageous for combined analysis of SAXS and SANS if the iANS give accurate data in q-range and also absolute intensity unit. For confirming it, we use 18 m SANS in HANARO and the results are shown in Fig. 1. 18 m SANS give wide and high quality data from 0.03 to 2 nm-1 only with the measuring time less than 1.5 h. Compare to it, the obtained profile from


51

iANS is only limitted q-range (0.2 <q< 6nm-1) but it exactly overlapped with the data from 18m SANS, indicating the accurate measurements in both q and absolute intensity dependence. Although the measurements time for iANS is more than 3 hours, the q-range (q>1nm-1) is more important for discussing the change in early stage of precipitates. Therefore, in the special case such as the characterization of precipitates and clusters with less than 3nm in diameter, iANS is quite competitive with the conventional SANS instruments.

2.2 Output of iANS from the view point of Materials Science The differential scattering SAXS and SANS profiles between samples tempered at certain temperature and without tempering shows scattering from the precipitates formed by tempering. The precipitates size for the sample tempered at 450ºC is about 1nm and those for 550ºC is about 3~5 nm. Interestingly, the intensity ratio between SAXS and SANS is about 20 and 10 for the samples tempered at 450ºC and 550ºC, respectively. This difference between two temperatures indicates that the composition of the formed precipitates are different, suggesting the formation of intermediate phase before forming stable phase. The results is analysed based on the recent proposal about forming ω phase in the martensite [4].

References [1] T. Shimizu, T. Koga, T. Noda, Denkiseiko 73 (2) (2002), 87–92 (in Japanese). [2] M. Ojima, M. Ohnuma, J. Suzuki, S. Ueta, S. Narita, T. Shimizu, Y. Tomota, Scripat Mater. 59 (2008) 313–316 [3] M. Ohnuma, J.Suzuki, S.Ohtsuka, S-W. Kim, T. Kaito, M.Inoue, H.Kitazawa, , Acta materialia, 57(2009), 5571-

5581 [4] D.H.Ping, Acta Metall. Sin., 28, 663–670(2015)

Fig 1 SANS profiles of subzero samples and tempered at 550ºC measured by 18m SANS (blue and red lines) and iANS (blue and red markers) in absolute unit.


52

Development of high temperature Titanium alloys, microstructure and property prediction methods

T. Kitashima1), Y. Yamabe-Mitarai1) 1) High Temperature Materials Unit, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan.

Phone: +81-29-859-2428, Fax: + 81-28-859-2501, Email: [email protected]

Abstract: High temperature materials are used in the compressor and/or the turbine of a jet engine. Increasing the temperature capability of these alloys leads to an improvement in the thermal efficiency of the jet engine, which results in a reduction of CO2 gas emission and fuel consumption. In order to design and develop new alloys which meet the requirement of engine makers and airlines, it’s necessary to consider not only one but several properties such as creep strength, yield strength, oxidation resistance, low-cycle fatigue, etc. However, it’s complicated and time-consuming to develop these alloys for practical use because they are multicomponent systems which contain six to ten elements. In this talk, we introduce the development of our high temperature titanium alloys for practical use, and a concept for one of future alloy design methods combining several models for microstructure and property predictions. In addition, we discuss the characteristics of recent commercial software to design high temperature alloys.

1. Alloy development of high temperature Ti alloys Global warming has been occurring at record level, and this has become one of the most pressing environmental issues in the world. The global warming is caused by the emission of greenhouse gases containing about 70% of carbon dioxide (CO2) which is partly produced by the global aviation industry. The amount of CO2 emission from this industry would increase in the future since more than 44,000 new aircrafts are expected to be introduced by 2036 [1]. On the other hand, airline companies are actively engaged in the reduction of CO2 emission by adopting a fuel efficient aircraft with high efficient jet engines, with a weight-reduced body, and with shape-improved body such as wing to reduce air resistance. According to a report by All Nippon Airways [2], the CO2 emission per seat for Boeing 787-8 which entered service in 2011 has been reduced by 20% compared to Boeing 767-300 which begun service in 1987 (Figure 1). In addition, the fuel consumption of Boeing 787-8 were suppressed by 15% of Boeing 767-300 [2], which was high on airline’s list of priorities from an economic perspective.

Figure 1 CO2 emission per seat for flights between Tokyo and Sapporo [2].

In order to increase the thermal efficiency of a jet engine, the weight-reduction of engine components by replacing to lighter materials, and the increase of gas temperature by using heat-resistant materials have been attempted. From these points of view, high temperature titanium alloy is attractive for use in jet engines because of their high strength-to-density ratio, especially this alloy has been used for high-pressure-compressor components such as compressor disks and blades. The maximum application temperature of titanium alloy has been increased as shown in Table 1. High-pressure compressor components in a jet engine are exposed to a severe environment with high temperature and high stress during flights. For example, these conditions cause the creep deformation of compressor blades and, in addition, these blades under centrifugal loading and vibration are damaged by fretting fatigue. In order to design and develop new alloys which meet the requirement of engine makers and airlines, it’s necessary to consider several properties such as creep strength, yield strength, oxidation resistance, fracture toughness, low-cycle fatigue, high-cycle fatigue. Titanium alloys for high temperature application basically consists of alpha and beta phases. The diffusion in alpha phase with hcp structure is much slower and deformation is more difficult than beta phase with bcc structure because of


53

limited number of slip systems. Therefore, the volume fraction of alpha phase in high temperature titanium alloy is high, e.g., about 95% at 800 oC in IMI 834. At high temperature beta phase is stable whereas at lower temperature than “beta transus temperature”, beta phase transforms to alpha phase.

Table 1 Maximum service temperature and chemical composition of conventional titanium alloys.

The composition of titanium alloy strongly affects the beta transus temperature. For example, Al, Ga, O, N which is called as alpha stabilizer increases the beta transus temperature. On the other hand, the addition of beta phase stabilizers such as V, Mo, Nb, Ta, decreases the beta transus temperature. The volume fraction of alpha phase can be determined by the lever rule as shown in pseudo binary phase diagram (Fig. 2 (a)). Figure 2 (b) shows a typical microstructure of near-alpha titanium alloy, Ti-6242, in where alpha phase appears as dark, and beta phase appears as bright phase. This microstructure consists of lamellar and equiaxed alpha structures. This lamellar structure is superior in creep strength, whereas equiaxed alpha structure increases fatigue strength. Fully lamellar structure is used for compressor blades. However, for compressor disks, the volume fraction of exuiaxed alpha phase is controlled to 15 to 30% as shown in Fig. 2 (b) by thermo-mechanical process conditions. Recent near-alpha alloys are strengthened by fine precipitates of alpha 2 phase with DO19 structure and silicides. However, during long exposure at high temperatures, these precipitates grow, coarsen, and cause embrittlement. The formation of alpha 2 phase is associated with the amounts of alpha stabilizing elements such as Al, Sn, Zr, Ga, O. Alpha Ti alloys tend to order when containing as little as 5 or 6 wt% Al and to embrittle forming Ti3Al particles when containing more than 8.5 wt% Al. The interaction between moving dislocations and alpha 2 particles, such as shearing particles or by-pass them, depends on the size, distribution, and the volume fraction of the particles. To predict the formation of alpha 2 phase, the empirical Al equivalent formula given by Al+1/2Ga+1/3Sn+1/6Zr+10O in wt% has been used. Figure 3 (a) shows the formation of alpha 2 phase in lamellar structure of a near-alpha Ti alloy with 10 of the Al equivalence. The effect of Al equivalence of near alpha alloy is shown in Fig. 4. The addition of Ga increases the Al equivalence increasing tensile strength because of solid solution strengthening and precipitation hardening due to the formation of alpha 2 phase, which, however, decreases ductility. The different amounts of Sn and Ga with an almost constant value but no more than 8.5 of the Al equivalence do not affect significantly the ductility at room temperature, where there was no precipitate of alpha 2 particles [3]. As a recommend final heat treatment temperature, IMI 834 is aged at 700 oC to precipitate the alpha 2 particles before service.

Figure 2 Schematic pseudo-binary phase diagram for alpha-beta titanium alloy and a typical microstructure of near-alpha titanium alloy Ti-6242.


54

Figure 3 Formation of alpha 2 particles and interaction of a germanide precipitate and dislocations. Figure 4 Effects of Si or Ge content and Al equivalent on tensile strength at 650 oC and elongation at room temperature. Another important strengthening is caused by the formation of silicides by Si addition. The solubility of Si in the matrix is low and large amounts of silicide formation causes embrittlement. Therefore, Si addition has been suppressed less than or equal to 0.45wt%. Even with the addition of small amount of Si in solution, the Si atoms interact with dislocations to increase the energy barriers for slip and cross slip increasing strength. In addition, the fine precipitates of the silicides retard dislocation. It has been reported the silicides in near-alpha alloys containing silicon and zirconium possesses the composition range of (TiZr)xSi where x may vary between 1.67 and 3 [4]. Also two types of hexagonal silicides, (TiZr)5Si3 and (TiZr)6Si3, have been found [5]. Recently, it has been reported Ge addition increases tensile strength by solid solution strengthening and precipitates of germanide [6]. Ge can form Ti5Ge3 germanide having the same crystal structure as Ti5Si3 silicide, and these intermetallics are completely in soluble in ternary Ti-Ge-Si system [7]. The lattice parameters a and c in Ti5(Si1-xGe)3 increase as Ge amount increases, in which the lattice misfit between the precipitate and alpha and beta phases could change, resulting in change of strain distribution near the interafce. The germanide addition and the simultaneous addition of Si and Ge in Ti-Al-Sn-Zr alloy increase the strength. Figure 3 shows a TEM image in the alloy showing the formation of intense slip bands and their interaction with a germanide precipitate. The addition of 1wt% Ge dissolving in alpha and beta phases increases tensile strength by solid solution strengthening. The additions of 4 wt% Ge or 1 wt% Si cause the formation of germanides or silisides which lower the ductility at room temperature (Fig. 4). This is because the fracture occurs due to the linkage of the cracks that are initiated in the regions where the slip bands intersect with precipitate particles. The tensile and creep strengths of near alpha alloys have been thus increased by the formations of alpha 2 phase, silicide and germanides although large amounts of those formations decrease ductility.

An exposure at high temperatures causes surface embrittlement of the compressor component due to oxide growth and oxygen penetration from surface, which degrades the mechanical properties due to the reduction of load-bearing thickness. This oxidation degradation becomes significant at higher temperatures than around 550C. Oxidation properties are dependent on microstructure (such as volume fraction of lamellar structure and the volume fraction of beta phase), partly because of the difference of diffusion rate of oxygen in grain and interior grain. During high temperature exposure of near-alpha Ti alloy, the formation and the growth of oxide layers including TiO2 and Al2O3 occur on the surface of substrate. The growth of these oxides is strongly affected by the diffusivity of Al, Ti, O ions in the scales. The external gas/oxide interface is always occupied by an Al2O3 layer in which the diffusivity of oxygen is smaller than TiO2. Therefore, in order to improve oxidation resistance, it’s necessary to enhance the formation of Al2O3 or to suppress the formation of TiO2. The oxidation kinetics of such near-alpha alloy is essentially dominated by TiO2 formation because the content of Al in conventional near-alpha titanium alloys is too low to form continuous protective Al2O3 layer on the surface, which is different from TiAl with protective Al2O3 layer.

The oxidation property is strongly dependent on the substrate composition because this determines the diffusion of ions in substrate and oxide scales, and aluminum activity at the TiO2/substrate interface for the formation of Al2O3. Figure 5 shows the mass gain per unit area of Ti-6242 and 1 wt% Si added Ti-6242, IMI834 which are oxidized at


55

750oC in air. The oxidation resistance of Ti-6242 is improved by Si addition forming SiO2 in oxide layers [8]. More recent alloy IMI 834 showed better oxidation resistance than Ti-6242 which is partly because of the beneficial effects of Si and Nb additions even though IMI 834 contains 4 wt% Sn which degrade oxidation resistance. During oxidation of Sn-bearing Ti alloy, Sn segregates at the TiO2/substrate interface as a metallic layer in the substrate which could reduce the aluminum activity at the interface suppressing the formation of Al2O3.

Figure 5 Mass gain vs time for Ti-6242, Ti-6242+1wt% Si, IMI 834 oxidized at 750oC in air.

To develop new alloys efficiently, it is necessary to understand the effect of composition in addition to

microstructure on the properties of Ti alloys. We have investigated the effect of Al, Sn, Ga, Zr, Mo, W, Ta, Nb, V, Si, Ge, and we have been building database for alloy development. A new Sn-free near-alpha Ti alloy could be one of possible future oxidation resistant alloys although Sn is a strong solid solution strengthener which is under investigation by the authors. One of our alloys have better creep strength and better oxidation resistance than conventional near-alpha alloy IMI 834 as shown in Fig. 6. Currently the fatigue properties of the alloy have been estimated and analyzed.

Figure 6 Our new alloy have better creep strength (a) and better oxidation resistance (b) than conventional near-alpha alloys.

3. Microstructure and property predictions One reason of the difficulty with prediction of the microstructure and properties of high temperature alloys is the complexity of the interaction of about 10 elements in the multicomponent systems. Therefore, to predict the microstructure and properties of multicomponent systems, it is necessary to prepare database which includes the effect of alloying. The accuracy for microstructure prediction has been gradually improved due to the recent development of thermodynamic database and mobility database in addition to the development of new physical models. However, it is still difficult to predict mechanical and environmental properties, with high accuracy, based on physical model-based calculations for high temperature alloys. We have been developing the models for microstructure evolution using a phase-field method [9, 10], creep deformation considering dislocation interaction and stacking fault energy [11], fatigue, and oxidation [12] for multicomponent alloys.

Our final target is to combine all those models, to include the interaction of those properties, and to predict the life time of components in jet engines in service based on alloy composition, thermo-mechanical processing condition and flight condition as shown in Fig. 7. A part of our concept was already reported [13, 14].


56

Figure 7 An example of a future alloy development tool. A part of this was introduced in Reference [14].

4. Conclusions High temperature titanium alloys have been developed for the application in jet engines. One of our new alloys have better creep strength and better oxidation resistance than conventional alloy, IMI 834. Such high temperature alloys are complicated multicomponent systems. In order to design new alloys, it is necessary to develop models and database for multicomponent systems. We have been developing the models for microstructure, creep deformation, fatigue, oxidation for high temperature alloys. Our final goal is to predict the life time of components in jet engines in service.

References [1] International Civil Aviation Organization. http://www.icao.int/

[2] All Nippon Airways. http://www.anahd.co.jp/en/

[3] T. Kitashima, K.S. Suresh, Y. Yamabe-Mitarai, S. Iwasaki, Mat. Sci. Forum, Vol. 783-786, (2014) 619.

[4] D.F. Neal, S.P. Fox, in: F.H. Froes, I.L. Caplan (Eds.), Titanium'92, Science and Technology, TMS, PA, 1993, pp.287–294.

[5] C. Ramachandra, A.K. Singh, G.M.K. Sarma, Metall. Trans., Vol. 24, (1993) 1273.

[6] T. Kitashima, K.S. Suresh, Y. Yamabe-Mitarai, Mat. Sci. Eng., A, Vol. 597, (2014) 212.

[7] T.A. Jain, C.G. Huang, C.E. Ho, C.R. Kao, Mater. Trans. JIM, Vol. 40, (1999) 307.

[8] T. Kitashima, Y. Yamabe-Mitarai, Metall. Mater. Trans. A, Vol. 46A (2015), 2758.

[9] T. Kitashima, H. Harada, Acta Materialia, Vol. 57, (2009), 2020.

[10] T. Ktiashima, Philosophical Magazine, Vol. 88, (2008), 1615.

[11] T. Ktiashima, Advanced Materials Research, Vol 1119, (2015), 580.

[12] T. Kitashima, L.Liu and H. Murakami, Journal of the Electrochemical Society, 160 (2013), C441.

[13] T. Kitashima, J. Wang H. Harada M. Sakamoto, T. Yokokawa, and M. Fukuda, Proceedings of International Gas Turbine Congress 2007, (2007), pp. 220.

[14] T. Kitashima, T. Yokokawa, M. Fukuda, H. Harada, Computer Simulations in Industry, A book in Japanese, (2010), pp. 9-26.


57

Overview of theme 3: Development of Multiscale Characterization in Structural Materials

Hideaki Kitazawa, Norimichi Watanabe, Hiroaki Mamiya National Institute for Materials Science, 1-2-1 Sengen, Tsukuba 305-0047, phone: +81-29-859-2818, fax: +81-298-2801, email:

[email protected]

Abstract: The theme 3 in SIP-IMASM has charge of development of multiscale characterization to investigate heterophase interface in structural materials. We introduce our research plans of theme 3 briefly. We have installed the high-temperature stage (up to 580 ºC) to our TOF-SIMS apparatus. The operand TOF-SIMS succeeded in detecting the solid-liquid coexistent phase in sputtered film of Al(25nm)/Co(2nm) on Si substrate.

1. Introduction

In a field of future’s commercial aircraft, the more active research and development for aircraft materials has been required to achieve enhancement of energy efficiency and reduction of CO2 emission. The project of SIP (Cross-ministerial Strategic Innovation Promotion Program) -SM4I (Structural Materials for Innovation) focuses on the crucial issues of R & D for the development of materials for aircraft engines, airframes and thermal power generation, (a) polymers and FRP (fiber reinforced plastics), (b) ceramics coatings, (c) heat resistant alloys and intermetallic compounds and (d) heat-resistant steel. The upgrading of high-specific strength properties and environmentally- and heat- resistant properties at higher temperatures are expected for (a) and for (b) and (c), respectively. The composite materials for (a) and (b) are adopted to achieve their improvement. The origin of fracture is estimated to be located in the vicinity of heterophase interface because of the difference of mechanical properties and thermal expansion under the real operation circumstance. The incompleteness of junction between different materials in the manufacturing process must have large effect on the lifetime and reliability of materials. On the other hand, the grain coarsening, segregation of impurities and degradation due to oxidization under the operating temperatures in (b), (c) and (d) must change the fine structures with time and give rise to the origin of fracture in weak points.

The Materials Integration hub is situated under the SIP-SM4I project. Main subjects of Materials Integration System are i) to predict the performance (life-time) of elements/structure which are manufactured from various choices of materials and processes, ii) to integrate theories, experimental knowledge, computation, measurement, database etc., and to utilize big-data, iii) contributing to reduce development time, to realize efficient development, to reduce manufacturing cost, to optimize the selection of materials and processes, to improve the reliability prediction and to reduce diagnosis and maintenance cost, and also iv) aiming to establish R&D center, capacity building and global network. The SIP-IMASM (Innovative Measurement and Analysis for Structural Materials) hub belongs to the Materials Integration hub in SIP- SM4I project. The SIP-IMASM hub must support the Materials Integration hub from experimental standpoints. The theme 3 in SIP-IMASM has charge of development of multiscale characterization to investigate heterophase interface in structural materials.

2. Research plans

The conventional microscopic observation methods, e. g. optical microscope, the scanning electron microscope (SEM) and the transmission electron microscope (TEM) have been utilized to detect the change of fine structures. Since the observation of fine structure was restricted in only the 2 dimensional information in a certain spatial range determined by each technique, it is not easy to evaluate the 3 dimensional distribution of grain boundary, nano-sized precipitation and complicated crack patterns in real materials with an enough spatial resolution and under the real operating circumstance by a single or a few uses of their techniques.

Fig. 1 Overview of investigation in theme 3.


58

In order to overcome the above mentioned issues, we have aimed at the following developments in theme 3 as shown in Fig. 1,

(1) Development of 3 dimensional imaging and fabrication techniques with multiscale characterization. (2) Development of the above techniques under the operating circumstances, e. g., high temperature or stress-strain field. The various experimental methods listed in the Table 1 will participate in R & D of the theme 3. We will focus on

the above two aims by the combination of these methods. In the next section, our recent experimental results will be presented by our developed operand TOF-SIMS.

Table 1. Experimental methods and their characteristic features in theme 3

Experimental method Characteristic features Target materials Focused Ion Beam-Scanning Electron Microscope (FIB -SEM)

3D observation in 10 mm X 10mm X 10mm with

1nm spatial resolution

Preparation of TEM sample

(b), (c) and (d)

3D atom probe (3D AP) Needle shape sample with 100nmin diameter

Observation of precipitation with a few – several tens

of precipitation

Applicability to insulators as well as metals

(b), (c) and (d)

Time-of-flight Secondary Ion Mass spectroscope (TOF-SIMS)

2D image with spatial resolution of 100nm Applicability to insulators as well as metals 3D image by sputtering Mass measuring range: 1 - 10,000 amu Operand measurements under high temperature (600

ºC)

(b), (c) and (d)

X-ray computed tomography (X-ray CT)

Nondestructive 3D imaging with spatial resolution of 1,000 nn to 700 nm Operand measurement under high temperatures and stress-strain fields

(a), (b) and (d)

Small angle X-ray scattering (SAXS)

Average evaluation of nano-size particles with 1 nm – 100nm Measurement spot size of 0.1mm Operand measurements under high temperature

(1,400 ºC) and stress-strain fields

(a), (b), (c) and

(d)

Multi-functional scanning probe microscope (SPM)

Spatial resolution of 1nm 2D distribution of elastic properties with topological image

(a) and (b)

Scanning helium ion microscope (SHIM

2D image with spatial resolution of 1nm Applicability to insulatorsOperand measurements under high temperature

(1,000 ºC)

Low damage to surface

(a), (b), (c) and

(d)

X-ray absorption fine structure - Computed tomography (XAFS-CT)

3D image with spatial resolution of 50 nm Operand measurement under high temperatures and stress-strain fields Element selectivity

(a), (b), (c) and

(d)

In-situ TEM Observation of structure change with atomic scale under high temperature (1,400 ºC) and stress-strain fields

(b), (c) and (d)


59

3. Resullts In TOF-SIMS, pulsed Ga

primary ions bombard the sample surface, and the mass spectrometry of atoms and /or molecules from the sample surface can be performed with high special resolution of 100 nm by measuring the time until the sputtered secondary ions reach a detector [1]. We have installed a high temperature stage in our TOF-SIMS apparatus (PHI TRIFT V nanoTOF, Ulvac-Phi Inc.). The temperature could reach 580 ºC. We tried to demonstrate the structure change of a test sample, the sputtered film Al(25nm)/Co(2nm) on Si substrate by this high temperature stage. The mass spectrum collection in the area of 100 μm x 100 μm of the sputtered film Al(25nm)/Co(2nm) by pulsed primary Ga ion beam and sputtering in the area of 300 μm x 300 μm by the Ar sputter gun were alternately carried out to construct 3D image. The Al layer at 580 ºC is completely different from the Al layer at 25 ºC as sown in Fig. 2. Since the eutectic point of Al-Si alloy is 575 ºC [2], a solid-liquid coexistent phase should exist when the temperature is above 575 ºC. Therefore, the complicated structure between Co and Si must be caused by the fact that the sample temperature is exceeded over the eutectic point. We will present other results in this symposium.

4. Conclusions We have installed the high-temperature stage (up to 580 ºC) to our TOF-SIMS apparatus. The operand TOF-SIMS

succeeded in detecting the solid-liquid coexistent phase in sputtered film of Al(25nm)/Co(2nm) on Si substrate.

Acknowledgement This work is supported by SIP (Cross-ministerial Strategic Innovation Promotion Program)-IMASM (Innovative

measurement and analysis for structural materials).

References [1] J. C. Vickerman and D. Briggs, “TOF-SIMS: Surface Analysis by Mass Spectrometry,” IM Publications,

Manchester, UK (2001).

[2] M.M. Makhlouf, H.V. Guthy, “The aluminum–silicon eutectic reaction: mechanisms and crystallography”, J. Light Met., 1(4), 199 (2002).

Fig. 2 3D-image of elements (Co, al and Si) in sputtered film of Al(25nm)/Co(2nm) on Si substrate at (a) 25 ºC and (b) 570 ºC In order to visualize inside of the Al layers, the tranceparancy of blue is changed from the right to the left pictures.


60

Microstructure characterization of structural materials using laser assisted 3D atom probe

K. Hono, T. Ohkubo and T. T. Sasaki National Institute for Materials Science, Tsukuba 305-0047, Japan

Abstract: Recent successful implementations of pulse lasers to assist field evaporation have expanded the application areas of the atom probe technique to a wide variety of materials including semiconductors and their thin film devices. The atom probe technique has been applied mainly to electrical conductive materials with a few successful reports of thin film oxides. In this talk, we will overview our work on the development of ultraviolet (UV) femtosecond (fs) laser atom probe to characterize nanoscale heterostructures in a wide variety of materials. All specimens were prepared from bulk or thin films by a combination of the lift-out technique and the annular focused ion beam (FIB) milling using a dual beam FIB machine. The employment of UV fs laser with the wavelengths of 343 and 258 nm made it possible to analyze bulk insulating ceramics routinely with excellent mass and spatial resolutions. In addition, it substantially improved the yield of success in atom probe analyses of "easily ruptured" specimens such as martensitic steels. The application of the UV laser not only made ceramics analysis possible, but also improved the mass resolution regardless of the thermal conductivity of specimens. This indicates that UV fs laser assisted atom probe can be a versatile technique to obtain atom tomography for all kind of inorganic martials. Based on our systematic work, we will discuss why short wave length is effective in improving the mass resolution and the yield of success of atom probe analyses. Then, we will present some of the recent results obtained from the structural materials, in particular the quantitative characterization of nanosized precipitates embedded in metal matrix.

References [1] K. Hono, T. Ohkubo, Y.M. Chen, M. Kodzuka, K. Oh-ishi, H. Sepehri-Amin, F. Li, T. Kinno, S. Tomiya, Y.

Kanitani, “Broadening the applications of the atom probe tecunique by ultraviolet femtosecond laser”, Ultramicroscopy, 111, 576-583 (2015)

[2] K. Hono and S. S. Babu, Physical Metallurgy, 5th Edition, eds by D. E. Laughlin and K. Hono, pp. 1453 – 1589, http://dx.doi.org/10.1016/B978-0-444-53770-6.00015-0.


61

Nanoscale Characterization of Structural Composite Materials

Daisuke Fujita1), Keiko Onishi 1), Hongxin Wang 1), Hideki Masuda 1)

1) National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, 305-0047, Japan

Tel +81-29-859-2741, Fax +81-29-859-2801 [email protected]

Abstract: Mechanical properties of structural composite materials shall be originated from the nanoscale and microscale characteristics including the multiscale structures and states of heterogeneous interfaces between the different phases. In order to respond to such demands, novel in situ characterization techniques with multiscale spatial resolution including nanoscale shall be developed. We have successfully developed such nanoscale characterization techniques. Low dimensional precipitates at the hetero interfaces on structural and functional materials have been characterized successfully by using an emerging high-resolution microscopy, or scanning helium ion microscopy (HIM). At first our recent developments of instrumentation for in-situ high resolution analysis at the elevated temperatures and the scanning transmission mode of HIM are introduced. Secondly the instrumental development of in situ mechanical property characterization using scanning probe microscopy (SPM) is introduced. Using a sharp probe made of bulk diamond, the true nanoscale hardness measurement by indentation, along with other mechanical and physical properties characterization, has been realized.

1. Introduction The mechanical and physical properties of advanced structural composite materials shall be closely related to the nanoscale characteristics of the structure, texture, and chemical and physical states at and near the interfaces between the different phases including nanoscale precipitates. The structural and physical properties of such low-dimensional nanoscale precipitates of structural and functional materials, including graphene and hexagonal boron nitride (h-BN) have been attracting the increasing attentions 1,2). In order to understand the mechanism of growth and the origins of high performance of mechanical and physical functions, both in situ and multi-functional characterization techniques with a high spatial resolution are highly demanded. We have been developing various techniques for such an active nano characterization using scanning probe microscopy (SPM), which can be operated in the various environments including high temperature, low temperature, magnetic field, light irradiation, external stress application, ultrahigh vacuum, controlled atmosphere, and so forth 3,4).

Such low dimensional nanoscale precipitates at the hetero interfaces on structural and functional materials can be characterized by a novel high spatial resolution microscopy, for example, scanning helium ion microscopy (HIM). At first our recent developments of instrumentation for in-situ high resolution analysis at the elevated temperatures and the scanning transmission mode of HIM shall be introduced here. Another important instrumental development is related to active nanocharacterization with atomic force microscopy (AFM) based SPMs. Since AFM can detect the interaction forces between the probe tips and the materials surfaces, it is intrinsically suitable or the characterization of mechanical properties of composite materials at the nanoscale. Using a sharp probe made of bulk diamond, the true nanoscale hardness measurement by indentation, along with other mechanical and physical properties characterization, has been realized.

2. Nanochacterization with Helium Ion Microscopy Although there are various methods to characterize such nanomaterials, novel imaging methods with a true nanoscale resolution have been demanded for realizing the controllable synthesis and optimization of structures, chemical states and textures. We have shown the instrumental developments of scanning helium ion microscope and its applications to nanomaterials 5). The schematic drawing of HIM is shown in Fig.1. Since it has a single-

Fig.1 Schematic of helium ion microscope using a single-atom-sized ion source.

Fig.2 Very high spatial resolution of 0.34 nm obtained for the edge of graphite by HIM.


62

atom-sized ion source formed on a mono-crystalline metal tip, high spatial-resolution and large depth-of-focus are realized. Another advantage is that HIM can observe insulating materials more easily than electron-beam based microscopies including SEM and TEM. The positive helium ion beam the A spatial resolution of 0.34 nm has been achieved at the edge of a graphite sample by SHIM as shown in Fig.2.

As for the instrumentation, we have developed a sample stage for transmission HIM (THIM) as shown in Fig. 3, which is suitable to analyze the internal structure of nanomaterials. An interesting application to hollow nanoparticles with HIM nanolithography will be shown. Another novel instrumentation is the development of in situ heating HIM as shown in Fig. 4. Interesting applications to temperature dependent change of C60 nano-whiskers will be introduced.

The 2-D h-BN nano-precipitates can be formed on the boron and nitrogen doped austenitic stainless steels through the surface and interfacial segregation and reaction of the boron and nitrogen dopants

2). We characterized the number and morphology of the h-BN nanosheets by using HIM (Fig.5). On the basis of the interaction between the scanning particles and h-BN nanosheets, we interpreted an exponential relationship between the intensities of images and the number of layers. Inelastic mean free paths of electrons and helium ions in h-BN nano sheets were calculated. Also quasi-free standing graphene nanoprecipitates on C-doped metallic substrates were characterized by HIM, along with Kelvin-probe force microscopy (KPFM), and scanning Auger-electron microscopy. The secondary electron imaging of HIM shows higher surface sensitivity and spatial

resolution than Rutherford backscattered He ion imaging, suitable for materials contrast imaging.

3. SPM-based nanocharacterization We have been developing the in situ characterization techniques at the nanoscale and micron scale for the high strength composite materials using SPM. The effectiveness of in situ nanoscale hardness characterization using diamond AFM probes is demonstrated on the composite materials including CFRPs and Ductiles.

4. Conclusion Nanopresipitates on structural materials have been characterized successfully using emerging high-resolution microscopy. Our recent developments on the in-situ analysis and the transmission mode for HIM and SPM, and comparison of different characterization shall be made in the presentation.

References [1] D. Fujita, “Nanoscale synthesis and characterization of graphene-based objects”, Sci. Technol. Adv. Mat. 12,

044611 (2011).

[2] M.S. Xu, D. Fujita, H.Z. Chen and N. Hanagata, “Formation of monolayer and few-layer hexagonal boron nitride nanosheets via surface segregation”, Nanoscale 3, 2854 (2011).

[3] D. Fujita and K. Sagisaka, “Active nanocharacterization of nanofunctional materials by scanning tunneling microscopy”, Sci. Technol. Adv. Mat., 9, 013003 (2008),

[4] N. Ishida, T. Iwasaki and D. Fujita, “Isotropic photo-decomposition of spherical organic polymers on rutile TiO2(110) surfaces”, Nanotechnology, 22, 155705 (2011).

[5] H.X. Guo, J.H. Gao, N. Ishida, M.S. Xu, D. Fujita, “Characterization of two-dimensional hexagonal boron nitride using scanning electron and scanning helium ion microscopy”, Appl. Phys. Lett. 104, 031607 (2014).

Fig.5 HIM image of h-BN nanoprecipitate on the

B- and N-doped auxtenitic stainless steel.

Fig.4 Development of high temperature SHIM.

Fig.3 Development of transmission helium ion microscopy (THIM).


63

Positron Annihilation Study of Vacancy-Type Defects in Iron and Steels

Y. Shirai

Department of Materials Science and Engineering, Kyoto University, Sakyo-ku Kyoto 606-8501, Japan Phone: +81-75-753-5466, Fax: +81-75-753-3579, E-mail: [email protected]

Abstract: Positron annihilation is the only method which can detect the size and the density of vacancy defects in metals quantitatively. This method has been successfully applied not only to functional materials but also to structural materials including iron and steels. Some of the important results obtained for the vacancy defects in iron and steels are reviewed after concise explanation of positron annihilation methods. On the other hand, positron annihilation method is non-destructive in principle. New positron-lifetime spectrometer has been developed in order to evaluate the damage and to predict the remaining-lifetime of structural materials including heat-resisting steels. Some of the typical results for fatigue-damaged or creep-damaged steels are shown.


64

Development of Performance Prediction System

Manabu Enoki1) 1) Department of Materials Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

TEL:+81-3-5841-7126 FAX:+81-3-5841-7181, [email protected]

Abstract: In our project of “Materials Integration” the performance prediction system for performances of welded structures is developed. The system consists of two groups of calculation modules and data base modules. Performances such as fatigue strength, creep strength, hydrogen embrittlement and brittle fracture are tried to be predicted by this system.

1. Introduction Many studies have been investigated to predict properties of materials using numerical calculation. Excellent results have been demonstrated when atomic structures directly contribute to the physical properties of materials. However, it is generally very difficult to predict the performances of materials which are necessary for structural materials because these macroscopic performances are induced by the complex relationship among many phenomena in different scales. Long period of time is usually needed to evaluate these performances in research and development of materials and a pace of development is limited by this time. The prediction of materials performances with some accuracy are expected very much in research and development of materials. On the other hand, many empirical rules have been also proposed to predict the performances of materials based on a certain theoretical consideration and accumulated huge experimental data of performances.

In our projects the development of performance prediction system for welded structures is planned to contribute the research and development of materials, where forward calculation modules for prediction of macroscopic performances such as fatigue strength, creep strength, hydrogen embrittlement, brittle fracture and so on will be developed using theoretical considerations and empirical rules and also data base modules for prediction will be prepared after the mining of enormous data of performances. The concept of our project is shown in Fig. 1. The prediction system for microstructure is differently developed and information of microstructure which is obtained from this system is used to predict time-dependent macroscopic performances of structural materials such as fatigue strength, creep strength, hydrogen embrittlement and brittle fracture. This performance prediction system consist of two groups of calculation modules, forward analysis modules based on theoretical physical models and data base modules using enormous accumulated performance data . Verification of system is also very important to obtain the prediction results with some accuracy and improve the system. Defects in materials with size distribution which are difficult to predict from the prediction system of microstructure are included into the modules of the prediction system of performances using both destructive and non-destructive experimental approaches. Time-dependent performances under certain conditions will be predicted by this system using materials and welding conditions even if microstructure information is not obtained.

2. Outline of System development 2.1 Development of calculation module Three-dimensional data of microstructure of materials due to material process conditions and geometry of structure is predicted by the microstructure prediction system and the group of calculation modules for performances in welded structure will predict the performance under service conditions based on the predicted microstructure data. Prediction of performances will be also done in some cases without the prediction of microstructure. Physical models are mainly used for the prediction of performances and the combination with data base modules based on the information statistical analysis will be able to enough consider the probabilistic or statistical scattering of data. Performances such as fatigue strength, creep strength, hydrogen embrittlement and brittle fracture are the objects to be predicted. The modules for fatigue strength prediction consist of two modules: one is the module with higher accuracy using three-dimensional microstructure information and Tanaka-Mura model for fatigue crack initiation and the other is the module with higher speed using statistically extracted parameters of microstructure and relatively simple geometry of microstructure and structure of samples. Predicted results by each module will be compared and modules for prediction will be improved or optimized. Creep deformation behaviours based on several types of constitutive laws are analysed and creep damage and fracture behaviours are calculated based on damage mechanics, fracture mechanics and diffusion theory in modules for creep strength prediction. Calculation modules include the consideration of multiple stress distribution which is very important for creep life prediction of thick structures and welded joints. Diffusional transportation of hydrogen in welded structures is calculated by the analysis of constitutive equation which controls the stress induced hydrogen diffusion in the prediction module for hydrogen embrittlement. Cracking due to hydrogen embrittlement is predicted by combining of these diffusional effects and three-dimensional stress distribution of welded structures. Inhomogeneous


65

microstructure of welded joints causes the dependence of diffusion phenomena on microstructure because of the effect of hydrogen trap elements. These effects are considered by referring the microstructural information predicted due to the microstructure prediction system. Module for the prediction of brittle fracture calculate the toughness of welded structures based on the difference of microstructure which is strongly controlled by melted metals and heat affected zone (HAZ) , and also the predicted results are combined with previously obtained results using Charpy impact test.

2.2 Development of data base module The performance data base module which can induce new empirical law or new data points is developed to improve the accuracy of prediction of materials performances by the arrangement and analysis of performance data which have been previously obtained. Organizations which join this project have enormous performance data regarding fatigue strength and creep strength and the empirical laws with certain accuracy which predict the performances of monolithic materials have been proposed. Data base module extended for welded structures is developed to improve the prediction of performances at welded joints. Improvement and restructuring of data base is planned using newly developed stochastic analysis tools and data mining techniques which are implemented in the data assimilation system. Complemental experiments for wanting data will be also done to upgrade the data base and the prediction system.

2.3 Verification of performance prediction system Performances of welded structures such as fatigue strength, creep strength, hydrogen embrittlement and brittle fracture are predicted by the above calculation and data base modules in prediction system for performances of materials. The predicted results are verified by both existing data and newly obtained data. Defects in materials which are difficult to obtain from the prediction system of microstructure are considered into the prediction system of performances using both destructive and non-destructive experimental approaches. Both calculation modules and data base modules for prediction of performances is improved by feedback of these verification results.

3. Remarks Many researches on the prediction of microstructure and properties of materials have been done using theory, empirical laws, numerical simulation and data base. However, only a few trials for integrated system of performance prediction of structural materials are done. Difficulties in prediction of performances for structural materials come from the complexity of time-dependent phenome and structures of materials in multi-scale levels. Recent development of computer engineering and information technology is successfully applied to the field of materials engineering, and hopefully this project lead the development of new generation structural materials and also enrich the international competitiveness of Japan in materials field.

Fig. 1. Concept of performance prediction system.


66

Overview of theme 4: Positron Annihilation Spectroscopy Based Research for the SIP-IMASM project

Nagayasu Oshima1) and SIP-IMASM thema-4 members 1) National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST),

Tsukuba, Ibaraki 305-8568, Japan. phone number: +81 29 861 5203, email address: [email protected]

Abstract: In the 4th theme in SIP-IMASM, we mainly study atomic-scale defects such as atomic defects and nano-voids in metals and intermolecular spaces (free volume) in polymers for developing high-performance structural materials. We mainly use positron annihilation spectroscopy (PAS) which is an effective method to analyse atomic-scale structure and defects. Examples of the research subjects in this 4th theme are briefly introduced.

1. Introduction

In general the macroscopic characteristics / properties of structural materials such as mechanical strength, heat-

resisting property and so on are affected by atomic-scale and/or nano-meter-scale structures of the materials. Atomic-scale and/or nano-meter-scale structures are sensitive to the environmental conditions and/or the external loading. Therefore, it is important to evaluate atomic-scale and/or nano-meter-scale structures when you develop high-performance structural materials and detect the early stage of deterioration of them.

There are four main researches themes (four groups) in the SIP-IMASM where various analytical methods for studying atomic-scale structure of high-performance structural materials are used. In our group (the 4th theme), we mainly study atomic-scale defects such as atomic defects and nano-void in metals/ ceramics and intermolecular spaces (free volume) in polymers. We mainly use positron annihilation spectroscopy (PAS) which is a powerful method to evaluate atomic-scale structure and defects and is applicable to various materials [1, 2, 3].

2. Positorn annihilation spectroscopy (PAS)

The positron is the anti-particle of an electron. When a positron encounters an electron, they annihilate into two gamma-rays with an energy of around E0 = 511 keV (This 511 keV is equivalent to the mass energy of the electron/positron). In the PAS, positrons obtained from radio-isotopes or accelerators are injected into material. Most of injected positrons into the material lose their energy and start thermal diffusion before the annihilation. Then, these thermalized positrons can effectively trap into the open-volume defects because there is no electrically repulsive force at the open-volume defects due to absence of positively charged nucleus. Positon annihilation parameters such as positron annihilation lifetime and energy-shift (Doppler-shift) of the annihilation gamma-ray from E0 is strongly dependent on the environmental structure at the positron annihilation site. For example positron annihilation time (i.e. positron lifetime) becomes longer if the size of the open volume defects become larger. Therefore, by measuring the positron lifetime spectrum and/or energy spectrum of annihilation gamma-ray we can evaluate atomic-scale defects, nano-voids, inter molecular spaces (free volume) etc. There are two main analytical methods in PAS, namely positron annihilation lifetime spectroscopy (PALS) and Doppler broadening of annihilation radiation (DBAR). We use three measurement techniques for PAS studies, i) fast positron method (bulk measurement methods), ii) slow positron beam method (near surface measurement method) and (iii) positron microbeam method (defect mapping measurement methods).

3. Resarch subjects using PAS

We plan to study several subjects by using PAS. Examples of subjects are listed in the bellow;

Development of non-destructive inspection method of structural materials applicable at the early stage of deteriorations. Study of relationship between free volume and macroscopic properties of polymers such as CRP. Study of defect-behaviour of heat-resistant-metals in the actual high-temperature environment. Study of revolution of points defects in the sample held in hydrogen embrittlement conditions. Study of defect distribution in structural materials (comparison before/after loading stresses) by using scanning positron microprobe.


67

Development of positron measurement system for realising the above subjects. For example, development of a positron beam manipulation method to improve beam energy resolution, development of a high-temperature sample holder for PAS apparatus and development of an automatic rapid-operation system for a scanning positron microprobe. Analysis/interpretation of the experimental data using first principle calculations.

Positron measurement techniques and progress in these subjects are presented in this conference.

Acknowledgement

This work was supported by the Cabinet Office strategic innovation creative program innovative structural materials (SIP).

References [1] https://unit.aist.go.jp/rima/xr-pos/group_detail/index_en.html [2] http://bukko.bk.tsukuba.ac.jp/~positron/index.html [3] http://pfwww.kek.jp/slowpos/


68

Vacancy-Type Defects and Open Spaces in Solid-State Materials Probed by Means of Positron Annihilation

Akira Uedono1 Nagayasu Oshima2, Ryoichi Suzuki2, and Shoji Ishibashi3 1Division of Applied Physics, Faculty of Pure and Applied Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan

2 Research Institute for Measurement and Analytical Instrumentation, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, 305-8568, Japan

3Nanomaterials Research Institute, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki 305-8568, Japan

Abstract: Positron annihilation is a powerful technique for evaluating vacancy-type defects and open spaces in solid-state materials such as metal, semiconductors, and polymers. In the present study, we report the application of positron annihilation to probe vacancies in plastically deformed GaN and free volumes (open spaces) in styrene-butadiene rubber (SBR) vulcanizates filled with carbon black (CB) or silica. For the GaN, vacancy-type defects were studied using monoenergetic positron beams. Measurements of Doppler broadening spectra of the annihilation radiation and positron lifetime spectra showed that both microvoids and Ga-vacancy-type defects were introduced into the deformed sample. The former defects were introduced through an agglomeration of vacancies introduced by dislocation motions. Lifetime spectra of positrons were measured for SBR vulcanizates with and without CB or silica. At temperatures between 10 and 420 K, no large difference between the size of the open spaces in the CB/SBR vulcanizate and that in the sample without the filler was observed. Above the glass transition temperature Tg (230 K), the same was true for the silica/SBR vulcanizate. Below Tg, however, the size of the open spaces was reduced by the incorporation of silica, due to the suppression of local molecular motions in the SBR. These studies show that the positron annihilation technique is a useful tool for an optimization of the fabrication process and design of industrial materials.

1. Introduction Positron annihilation is a powerful technique for evaluating vacancy-type defects and open spaces in solid-state materials. Because there is no restriction on sample conductivity, it can be used for metals, semiconductors, and polymers, etc. For crystalline materials, detectable defects are monovacancies to vacancy clusters. For amorphous polymers, sizes of open spaces (free volumes) and their densities can be estimated from the measurements of lifetimes of positronium (Ps) atoms and their annihilation rate. Because there is no restriction on sample temperature, vacancies and free volumes can be studied at elevated temperatures. In conventional positron annihilation experiments, high-energy positrons (≤0.54 MeV) emitted from 22Na are implanted into the samples. Because the maximum penetration depth of such positrons is in the order of 0.1 mm, one can obtain information about the bulk properties. By using slow/monoenergetic positron beams, implantation profiles of monoenergetic positrons can be adjusted to the region of interest in the sample by accelerating the positrons to the desired energy. In general, the incident energy of monoenergetic positrons varies from several eV to 30-50 keV. Therefore, the regions sampled by such positrons can vary from those on the surface of a specimen to those at a depth in the order of μm. In the present study, we report the application of the positron annihilation to probe vacancies in plastically deformed GaN and free volumes in styrene-butadiene rubber (SBR) vulcanizates filled with carbon black (CB) or silica.

2. Analysis methods When a positron is implanted into condensed matter, it annihilates with an electron and emits two 511-keV γ quanta [1,2]. Scheme of positron annihilation experiments is shown in Fig. 1. The energy distribution of the annihilation γ rays is broadened by the momentum component of the annihilating electron-positron pair pL, which is parallel to the emitting direction of the γ rays. The energy of the γ rays is given by Eγ = 511 ± ΔEγ keV. Here, the Doppler shift ΔEγ is given by ΔEγ = pLc/2, where c is the speed of light. A freely diffusing positron may be localized in a vacancy-type defect because of Coulomb repulsion from positively charged ion cores. Because the momentum distribution of the electrons in such defect differs from that of electrons in the bulk material, these defects can be detected by measuring the Doppler broadening spectra of the annihilation radiation. The lifetime of positrons trapped by vacancy-type defects increases because of the reduced electron density in such defects. Information obtained by measuring the lifetime spectra of positrons is useful for identifying vacancy-type defects. Figure 2(a) and (b) show Doppler broadening spectrum and the lifetime spectrum of positron for GaN, respectively. The solid curves shown in Fig. 2(a) are the results obtained by the first principles calculation [3]. The changes in the spectra due to the trapping of positrons by vacancies are characterized by the S parameter, which mainly reflects changes due to the annihilation of positron-electron pairs with a low-momentum distribution, and by the W parameter, which mainly characterizes changes due to the annihilation of pairs with a high-momentum distribution (Fig. 2).


69

For amorphous materials, positronium (Ps: a hydrogen-like bound state between a positron and an electron) may form in open spaces [4]. Ps exhibits two spin states: para-Ps (p-Ps) is a singlet state and ortho-Ps (o-Ps) is a triplet state. The intrinsic lifetimes of p-Ps and o-Ps are 0.125 ns and 142 ns, respectively. Ortho-Ps primarily exhibits three-photon (3γ) annihilation that produces a continuous energy distribution from 0 to 511 keV. Because p-Ps decays from 2-γ process, the energy of such γ-rays is 511 keV. When o-Ps is trapped by open volumes, the positron involved in o-Ps may annihilate with an electron of pore interiors to emit two γ rays before 3γ-annihilation (pick-off annihilation). A large open spaces decreases the probability of this process and increases the o-Ps lifetime. Thus, one can estimate the size of open volumes from measurements of the o-Ps lifetime.

Fig. 1. Scheme of positron annihilation experiments. Positrons from an isotope source penetrate the sample. After thermalization of energetic positrons, they diffuse in the sample before the annihilation.

Fig. 2. (a) Doppler broadening spectrum of the annihilation γ-rays and (b) the lifetime spectrum of positron for GaN.

The solid curves shown in Fig. 2(a) are the results obtained by the first principles calculation.

Fig. 3. Schematic drawing of Doppler broadening spectrum for (a) the annihilation of positrons from delocalized state. The spectra for annihilation of positrons trapped by (b) a monovacancy and (c) a vacancy-cluster are also shown. The Doppler broadening spectrum is characterized by the S parameters, which mainly reflects changes due to the annihilation of positron-electron pairs with a low-momentum distribution, and by the W parameter, which mainly characterizes changes due to the annihilation of pairs with a high-momentum distribution.

2. Experiment The samples investigated were oxygen-doped GaN grown by hydride vapor phase epitaxy (HVPE) using the lateral epitaxial overgrowth (LEO) technique, where the crystal grew over SiO2 masks on GaAs {111} substrates [5]. Samples with a rectangular shape (1.0 ⋅ 1.0 ⋅ 3.6 mm) were cut from the crystal by a dicing machine. Compressive stress was applied to the sample in Ar-gas flowing atmosphere at 950°C using an Instron-type machine, where the stress axis was inclined at 45˚ with respect to the [11 2 0] direction. The deformation was applied up to the shear strain of ε = 4%. The


70

typical dislocation density was 109 cm−2 in similarly deformed samples. The Burgers vector b of the major dislocations

was b= (1/3)[11 2 0] suggesting they were pure edge dislocations. We refer to this sample as ‘controlled’. With a monoenergetic positron beam, the Doppler broadening spectra of the annihilation radiation were measured with a Ge detector as a function of the incident positron energy E [6]. The lifetime spectrum of positrons was measured using a pulsed monoenergetic positron beam. A brightness enhancement method was used to reduce the beam size at the sample to about 0.1 mm. Details of the positron focusing system are described elsewhere [7]. Positron annihilation Doppler broadening spectra and positron lifetimes were calculated with our computational code QMAS [3].

We used CB- and silica-loaded SBR vulcanizates [8]. The SBR contains 15% bound styrene and 57% vinyl (solution SBR: SL574 from JSR Co.). The CB had a mean particle size of 27.5 nm (Showa Cabot, ASTM No. N330), and the silica had a mean particle size of 30 nm (Nippon Silica Industrial, Nipsil AQ). A one-to-one mixture of organosilane and CB (Degussa, X50S) was used as the coupling agent in the silica/SBR vulcanizate. For SBR with CR and Silica, the phr values of CBR and CR were 43.2 and 40, respectively. We measured the lifetime spectra of positrons at temperatures between 10 and 420 K in vacuum at 10−

4 Pa. The lifetime spectrum of the positrons is expressed as )λexp(Σλ)( ii tItN i−= , where iλ and iI are the annihilation rate of positrons of the i-th component and its intensity,

respectively. The lifetime of positrons for the i-th annihilation mode iτ is given by i1/λ .

3. Results and discussions 3.1 Vacancy-type defects introduced by plastic deformation of GaN Figure 4 shows the S values of the samples before and after deformation as a function of incident positron energy E. The S−E curve for an undoped sample grown by HVPE is also shown, and the result was the typical one for defect free GaN [5,9]. For the controlled samples, the S value at E = 3-6 keV were almost constant, suggesting the presence of vacancy-type defects in the subsurface region. The defect rich region is considered to be introduced by the cutting process of the sample. For the deformed sample, the observed increase in the S value can be attributed to the vacancy-type defects introduced by the deformation.

Figure 5 shows the mean positron lifetime τM as a function of the sample position for the controlled and deformed samples. For the deformed sample, the obtained distribution of the τM value can be attributed to the distribution of vacancy-type defects introduced during the deformation. For the controlled sample, the τM values tend to increase near the edge of the sample, suggesting the trapping of positrons by defects introduced during the sample-cutting process. The lifetime spectrum measured at the center of the deformed sample was decomposed into two components, and the obtained τ1, τ2, and I2 were 222 ± 2 ps, 456 ± 6 ps, and 15 ± 1%, respectively. The spectrum for the controlled sample can be analyzed assuming one component, and the obtained lifetime was 183 ± 2 ps. Using the QMAS code, the positron lifetime in defect-free (DF) GaN was calculated to be 159 ps. The ratio of the lifetimes trapped by the vacancy to the DF value is, for example, 1.49 for VGa, 1.54 for VGaVN, and 2.60 for (VGaVN)6. For the deformed GaN, the ratios of τ 1 and τ2 to the positron lifetime in the undoped sample (150 ps) were 1.48 and 3.04, respectively. Thus, the second components for the deformed sample can be attributed to the trapping of positrons by large vacancy clusters (n > 6) or microvoids. When the deformation was done at a temperature higher than the annealing temperature for vacancy-type defects, they could interact with each other and form vacancy or interstitial clusters. Thus, the observed microvoids are considered to be introduced by the above mechanism.

Fig. 4. S parameters as a function of incident positron energy E for the controlled, and deformed samples. The S−E

curve for the undoped sample is also shown.


71

Figure 6 shows the (S,W) values for the undoped, controlled, and deformed samples. The simulated values corresponding to the annihilation of positrons in the defect free (DF) state and those of positrons trapped by VGa, (VGaVN)n (n=1, 2, 3, and 6) and VGa(ON)4 are also shown. In the S−W plot, if the major trapping centre of positrons is VGa, we would expect to find the (S,W) values on the lines which connects the (S,W) values for DF and VGa [5,9]. For the deformed sample, the dominant annihilation mode is the first component (85%) which corresponds to the annihilation of positrons trapped by VGa-type defects. The observed (S,W) value, however, is located rather far from the line connecting the DF values and VGa or VGaVN. The (S,W) value for (VGaVN)n tends to shift toward the lower right with increasing n. Thus, the observed deviation can be attributed to an effect of the annihilation of positrons trapped by microvoids. For the controlled sample, the (S,W) values measured at E = 4 and 30 keV and the value for the undoped sample lie on a line, suggesting that the positron implanted with E = 30 keV (the mean implantation depth corresponding to this energy is about 2 μm) could be annihilated in the trapped state by a similar defect species near the subsurface. For this sample, the (S,W) value at E = 4 keV is located downward from the value for the deformed sample. The vacancy-type defects are considered to be introduced during the sample cutting process by a similar characteristic of the deformation. However, because the dicing/polishing processes were done at room temperature, the formation of microvoids could be suppressed, and the effect of the positron annihilation of such defects is therefore small.

Fig. 5. Distribution of the mean positron lifetime for the controlled and as-grown samples. An inset shows the sample

geometry, where a horizontal arrow shows the compression axis.

Fig. 6. (S,W) values (brown symbol) for the undoped, controlled, and deformed samples. The incident energy of

positrons is shown in the figure (E = 4 or 30 keV). The (S,W) values (blue symbol) calculated by QMAS for positron annihilation in defect-free (DF) and typical vacancy-type defects are also shown.

3.2 Free volumes in Carbon-Black- and Silica- Loaded SBR The lifetime spectra of positrons for SBR vulcanizate were decomposed into three components. The lifetime spectra for the fillers were measured, and subtracted them from the results obtained for the samples with those fillers. Figures 7 and 8 show the temperature dependencies of the lifetime and intensities corresponding to the pick-off annihilation of o-Ps (τ3 and I3) for the silica/SBR, CB/SBR vulcanizates, and the sample without filler, respectively. For the sample without filer, using the model to estimate the free volume size [4] and the obtained value of τ3, the size of the open spaces was estimated to be 0.038 nm3 at 10 K and 0.19 nm3 at 420 K. The onset temperature of the change in the temperature


72

gradient of τ3 and I3 can be connected to that of the relaxation processes of molecular motions. In the temperature range between the grass transition temperature Tg and 310 K, the temperature gradient of τ3 was larger than that below Tg. This was due to the cooperative motion of large segments of molecules. Above 310 K, however, the increase in τ3 became saturated. This saturation may have been due to the suppression of molecular motions by the network structure of SBR vulcanizates. In Fig. 10, above Tg, the value of I3 increased with temperature. This is due to the increase in the density of the open spaces due to the successive increase in the mobility of molecules. The increase in I3 below Tg can be interpreted as an increase in the density of spur electrons [4] available for Ps formation. Those electrons are considered to be loosely trapped by radicals introduced by the irradiation of positrons or by the polarization forces of the molecules; the shallow traps become effective after the freezing of the molecular motions of the main chain. In the present experiments, therefore, the behavior of I3 below Tg can be attributed to such an origin.

As shown in Fig. 7, the value of τ3 for the CB/SBR vulcanizate agrees with that for the sample without filler. Thus, it can be concluded that incorporating CB into SBR did not affect the size of the open spaces. The τ3 values for the silica/SBR vulcanizate above Tg were close to those for the sample without filler, but below Tg, they were not. The observed decrease in the size of the open spaces can be attributed to the suppression of local molecular motions by the incorporation of silica; the ratio of the volume of the open spaces in the silica/SBR vulcanizate to that in the specimen without filler was approximately 0.6 at 10 K. Above Tg, the value of I3 for the specimens with filler was smaller than that for the specimen without filler (Fig. 8). This can be attributed to the decrease in the density of the open spaces caused by the incorporation of filler. Above 390 K, as mentioned above, the value of I3 for the sample without filler started to increase. For the CB/SBR vulcanizate, however, the increase of I3 is suppressed. For the silica/SBR vulcanizate, the value of I3 above 400 K is larger than that for the CB/SBR vulcanizate. Thus, the suppression of molecular motions by silica is likely to be diminished at high temperature.

τ

Fig. 7. Temperature dependence of τ3 for SBR with and without filler.

Fig. 8. Temperature dependence of I3 for SBR with and without filler. The dip in the dependence at 230 K is attributed

to the onset of cooperative motion of large segments of molecules.


73

Below Tg, the increase in I3 corresponding to the enhancement of Ps formation was suppressed in the CB/SBR vulcanizate. It is probable that the functional groups at the surface of CB (carbonyl, carboxyl, etc.) act as electron traps, inhibiting Ps formation. However, since no large difference between the values of I3 for the specimens with filler was observed above Tg, the suppression of Ps formation by CB is probably small above Tg.

4. Conclusion We studied vacancy-type defects in plastically deformed GaN using monoenergetic positron beams. Doppler broadening spectra were measured as a function of the incident energy of positrons for the samples before and after the deformation. Before the deformation, a vacancy-rich region was observed in the subsurface region (0-150 nm), which was attributed to the defects introduced during dicing and/or polishing of the samples. For the deformed sample, the observed lifetime spectrum was decomposed into two components. The second lifetime τ2 was derived to be 456 ps, and this annihilation mode was attributed to the annihilation of positrons trapped by microvoids. Such defects were considered to be introduced by reactions between vacancies introduced by dislocation motions. We have shown that the positron annihilation parameter is sensitive to vacancy-type defects introduced by dislocation motions, meaning that this technique can be a useful tool for studying device degradation that is related to dislocations.

We also investigated open spaces and the molecular motions of CB- and silica-loaded SBR vulcanizates. The lifetime spectra of positrons were measured for specimens with and without filler, and the temperature dependencies of the lifetime of o-Ps and its intensity were determined. For the CB/ and silica/SBR vulcanizates, the parameters corresponding to the annihilation of o-Ps in SBR were derived using the results obtained for the CB and silica themselves. At temperatures between 10 and 420 K, no large difference between the size of the open spaces in the CB/SBR vulcanizate and that in the specimen without filler was observed. Thus, the incorporation of CB does not affect the size of the open spaces. Above Tg (230 K), the same was true for the silica/SBR vulcanizate, but below Tg, the size of the open spaces was decreased by the incorporation of silica. This is attributed to the suppression of local molecular motions of SBR. The density of the open spaces was decreased by the incorporation of filler, but above 400 K, it started to increase in the silica/SBR vulcanizate. For the CB/SBR vulcanizate, the introduction of the open spaces was suppressed, even at 420 K. Below Tg, Ps formation was enhanced for the specimen without filler, but was suppressed in the CB/SBR vulcanizate. This is attributed to the inhibition of Ps formation by the functional groups at the surface of CB. Our investigation has demonstrated the potential of using positron annihilation techniques to study the open spaces and molecular motions in SBR vulcanizates designed for automotive tires.

We have shown that positron annihilation parameters are sensitive to vacancy-type defects introduced by the mechanical deformation of materials and that dynamics of free volumes in amorphous polymer. These studies suggest that the positron annihilation technique is a useful tool for an optimization of the fabrication process and design of industrial materials.

References [1] P. Hautojärvi, Positron in Solids, Topics in Current Physics 12 (Springer-Verlag, Berlin, 1979).

[2] R. Krause-Rehberg and H. S. Leipner, Positron Annihilation in Semiconductors, Solid-State Sciences 127 (Springer-Verlag, Berlin, 1999).

[3] S. Ishibashi and A. Uedono, “First-principles calculation of positron states and annihilation parameters for group-III nitrides”, J. Phys.: Conf. Ser. 505, 012010 (2014).

[4] Y. C. Jean, P. E. Mallon, R. Zhang, H. Chen, Y. C. Wu, Y. Li, J. Zhang, Principle and Application of Positron and Positronium Chemistry (World Scientific, 2003) pp. 281.

[5] A. Uedono, I. Yonenaga, T. Watanabe, S. Kimura, N. Oshima, R. Suzuki, S. Ishibashi, and Y. Ohno, “Vacancy-type defects introduced by plastic deformation of GaN studied using monoenergetic positron beams”, J. Appl. Phys. 114, 084506 (2013).

[6] A. Uedono, S. Ishibashi, N. Oshima, and, R. Suzuki, “Positron annihilation spectroscopy on nitride-based semiconductors”, Jpn. J. Appl. Phys. 52, 08JJ02 (2013).

[7] N. Oshima, R. Suzuki, T. Ohdaira, A. Kinomura, T. Narumi, A. Uedono, and M. Fujinami, “Rapid three-dimensional imaging of defect distributions using a high-intensity positron microbeam”, Appl. Phys. Lett. 94, 194104 (2009).

[8] A. Uedono, S. Fukui, M. Muramatsu, T. Ubukata, S. Kimura, and S. Tanigawa, “Open spaces and molecular motions in carbon-black- and silica-loaded SBR investigated using positron annihilation”, J. Polym. Sci. B 39, 835 (2001).

[9] A. Uedono, S. Ishibashi, K. Tenjinbayashi, T. Tsutsui, K. Nakahara, D. Takamizu, and S. F. Chichibu, “Defect characterization in Mg-doped GaN studied using a monoenergetic positron beam”, J. Appl. Phys. 111, 014508 (2012).


74

Performance test of a pulse stretch system for materials science at KEK Slow Positron Facility

Ken Wada1), Masaki Maekawa2), Izumi Mochizuki3), Masao Kimura3), Toshio Hyodo3) 1) Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK),

1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan phone number: +81 29 864 5661, email address: [email protected]

2) Quantum Beam Science Center, Sector of Nuclear Science Research, Japan Atomic Energy Agency, 1233 Watanuki, Takasaki, Gunma 370-1292, Japan

3) Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan

Abstract: A performance test of a pulse stretch system for a linac-based slow-positron beam is briefly reported. A part of positrons of 1-μs pulse width generated by using a linac operated at 50Hz are confirmed to be stretched up to 20 ms with an energy of 2.5 keV.

1. Introduction The Slow Positron Facility (SPF) at the Institute of Materials Structure Science (IMSS), High Energy Accelerator Research Organization (KEK) provides a high-intensity pulsed slow-positron beam by using a dedicated linac operated at 55 MeV, 0.6 kW [1, 2]. Accelerated electrons are impinged on a Tantalum (Ta) converter and deflected in the electric field of a high-Z atom, emitting Bremsstrahlung X-rays. The radiation causes positron-electron pair creation in the same metal. The energy range of the generated positrons are very wide up to almost 55 MeV. A moderator, which is composed of 25-μm Tungsten (W) films, is used to obtain a mono-energetic positrons (slow positrons). When high-energy positrons are injected on a metal, the positrons are thermalized in the bulk and most of them annihilate with an electron. But in particular metals, like W, Ni, and Cu, a part of positrons are reemitted with a certain energy after thermalization. These metals have negative work function for the positron, and thus emit positrons with the corresponding energy. The moderated positrons are then accelerated to 0.1 - 35 keV and transported by magnetic field to experiment stations. This slow-positron generating process occurs within ps order, resulting in that the slow-positron beam reflects the time structure of the accelerated electron beam with ns- to μs-pulse width from the linac.

The present experiment stations at KEK Slow Positron Facility are performed with a pulsed slow-positron beam directly. There are three experiment stations, namely, positronium negative (Ps-) ion station [3-6], positronium time-of-flight (Ps-TOF) station [7, 8], and total-reflection high-energy positron diffraction (TRHEPD) station with a brightness enhancement system [9-12]. Photodetachment of the Ps- ions produced by using the pulsed positron beam has been accomplished by irradiation of 25Hz pulsed (12-ns width) Q-switched Nd:YAG laser. The KEK slow-positron beam operated with a 50Hz short-pulse (1-10 ns) mode fits this experiment. The data with and without the laser irradiation are obtained simultaneously. This short-pulse beam is also used for the Ps-TOF experiment with sufficient time resolution. For TRHEPD experiment, a high-intensity long-pulse (∼1 μs) beam is provided because the TRHEPD apparatus does not depend on the time structure of the pulse beam and prefer higher intensity.

However, conventional positron annihilation experiments such as positron annihilation lifetime spectroscopy (PALS) and Doppler broadening spectroscopy (DBS) cannot take advantage of the high-intensity beam with a pulsed one because of the pileup problem at the detectors. Since those positron annihilation experiments are effective tools for structural materials science, we have been developing a pulse stretch system with an energy of 5 keV, which allows full use of high-intensity pulsed beam for the annihilation experiments.

2. Pulse stretch system for slow-positron beam We have introduced a cylindrical entrance electrode, a 6-m long trap electrode, and an exit electrode for pulse stretching. The entrance electrode voltage set at 5.1 kV is changed to be 4.6 kV when a 1-μs pulsed positron beam with an energy of 4.8±0.05 keV are reached the entrance electrode to let the beam in a linear storage section. Positrons then goes to the exit electrode at 5.0 kV and reflected backword. Before the positrons travelled back to the entrance electrode, the entrance electrode voltage is changed to be 5.1 kV and thus positrons are trapped in the linear storage section. The trap electrode voltage is then increased gradually, letting the positrons go through the exit electrode at 5 keV in order of the kinetic energy. By adjusting the sweeping speed of the trapping voltage, we are able to obtain a stretched pulse beam up to 20 ms width. This cycle is operated at 50Hz synchronized with the linac.

There is a pulse stretch system operated at National Institute of Advanced Industrial Science and Technology (AIST) for a ∼10 eV beam [13] based on another system. Instead of varying the trap electrode voltage, the exit electrode voltage is varied to let the trapped positron go through it.


75

With our new pulse stretch system, we are planning to obtain 5-keV pulse beam stretched with a narrow energy dispersion. This system would be effectively applied for high efficiency PALS experiments with a chopper, pre-buncher and RF buncher in order to produce much shorter pulses for PALS measurements [14]; it is important to provide positrons with a narrow energy dispersion to bunchers for the efficiency. A pulse stretched beam with an energy of 5 keV is also effectively used with a transmission type remoderator with the grounded potential without a high-voltage station for the brightness enhancement system; positrons should be injected into the transmission type remoderator at around 5keV kinetic energy.

3. Experiment We have conducted a performance test of a newly introduced pulse stretch section with 1-μs pulse beam with an energy of 2.5 keV. A gate valve downstream of the pulse stretch section is closed to detect γ-rays emitted from annihilating pulse-stretched positrons at the valve. A plastic scintillator mounted on a photomultiplier tube (PMT) is used to detect the γ-rays, and the anode signal from the PMT is observed by a digital oscilloscope (Fig 1).

Fig. 1 A voltage variation of the trap electrode of the pulse stretch section and an anode signal corresponding to the γ-rays from pulse-stretched positrons annihilating at a gate valve downstream of the pulse stretch section.

The anode signal are observed to be corresponding to the change of the trap-electrode voltage. A 1-μs pulse beam has been stretched to 15 ms at every 20 ms (50Hz) in Fig 1. When modulating the time variation of the trap-electrode voltage, the anode signal width varies corresponding to it up to almost 20 ms, which is the upper limit of the 50Hz repetition of the linac.

Annihilation signal peaks are also observed at every prompt of 20 ms cycle. A calculation for the spatial voltage distribution around the present exit electrode shows that a part of positrons was not able to be trapped and goes through the electrodes and hit directly on the gate valve. We are working on improving the electrodes to trap the positrons more efficiently.

4. Conclusion A pulse stretch system for a linac-based slow-positron beam are introduced to the KEK Slow Positron Facility. The system has been tested and confirmed that we are able to stretch 1-μs pulses up to 20 ms with an energy of 2.5 keV at 50Hz synchronized with the linac. A part of the positrons, however, could not be trapped at the pulse stretch section. We are improving the electrodes to trap the positrons more efficiently at the pulse stretch section.

Acknowledgement The present work was partly supported by Toray Science and Technology Grant from Toray Science Foundation, a Grant in-Aid for Scientific Research, Grant No. (S) 24221007, from JSPSA, and Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.


76

Reference [1] K. Wada, T. Hyodo, A. Yagishita, M. Ikeda, S. Ohsawa, T. Shidara, K. Michishio, T. Tachibana, Y. Nagashima, Y.

Fukaya, M. Maekawa, and A. Kawasuso, Eur. Phys. J. D, 66, 37-1-4 (2012).

[2] K. Wada, T. Hyodo, T. Kosuge, Y. Saito, M. Ikeda, S. Ohsawa, T. Shidara, K. Michishio, T. Tachibana, H. Terabe, Y. Nagashima, Y. Fukaya, M. Maekawa, I. Mochizuki, and A. Kawasuso, J. Phys.: COnf. Ser. 443, 012082-1-6 (2013).

[3] T. Tachibana, K. Michishio, H. Terabe, K. Wada, T. Hyodo, T. Kurihara, A. Yagishita, and Y. Nagashima Nucl. Instrum. Methods Phys. Res., Sect. A, 621, 670-672 (2010).

[4] K. Michishio, T. Tachibana, H. Terabe, K. Wada, T. Kuga, A. Yagishita, T. Hyodo, and Y. Nagashima, Phys. Rev. Lett., 106 153401-1-4 (2011).

[5] K. Michishio, T. Tachibana, R. H. Suzuki, K. Wada, A. Yagishita, T. Hyodo, and Y. Nagashima, Appl. Phys. Lett., 100, 254102-1-4 (2012).

[6] K. Michishio, R. H. Suzuki, K. Wada, I. Mochizuki, T. Hyodo, A. Yagishita, Y. Nagashima, Nucl. Instrum. Methods Phys. Res., Sect. A, 785, 5-8 (2015).

[7] H. Terabe, S. Iida, K. Wada, T. Hyodo, A. Yagishita, and Y. Nagashima, J. Phys.: Conf. Ser., 443, 012075-1-4 (2013).

[8] H. Terabe, S. Iida, T. Yamashita, T. Tachibana, B. Barbiellini, K. Wada, I. Mochizuki, A. Yagishita, T. Hyodo, and Y. Nagashima, Surf. Sci., 641, 68-71 (2015).

[9] I. Mochizuki, Y. Fukaya, A. Kawasuso, K. Yaji, A. Harasawa, I. matsuda, K. Wada, and T. Hyodo, Phys. Rev. B, 85, 245438-1-6 (2012).

[10] Y. Fukaya, I. Mochizuki, M. Maekawa, K. Wada, T. Hyodo, I. Matsuda, and A. Kawasuso, Phys. Rev. B, 88, 205413-1-4 (2013).

[11] Y. Fukaya, M. Maekawa, A. Kawasuso,I. Mochizuki, K. Wada, T. Shidara, A. Ichimiya, and T. Hyodo, Appl. Phys. Express, 7, 056601-1-4 (2014).

[12] M. Maekawa, K. Wada, Y. Fukaya, A. Kawasuso, I. Mochizuki, T. Shidara, and T. Hyodo, Eur. Phys. J. D, 68, 165-1-6 (2014).

[13] T. Akahane, T. Chiba, N. Shiotani, S. Tanigawa, T. Mikado, R. Suzuki, M. Chiwaki, T. Yamazaki, and T. Tomimasu, Appl. Phys. A, 51, 146-150 (1990).

[14] N. Oshima, R. Suzuki, T. Ohdaira, A. Kinomura, T. Narumi, A. Uedono, and M. Fujinami, Appl. Phys. Lett. 94, 194104-1-3 (2009).


77

Poster papers

IMASM Theme 1 Stress and Cracks


78

Development of X-ray Multiple Image Radiography at the Photon Factory

Keiichi Hirano, Kazuyuki Hyodo, Yumiko Takahashi and Masao Kimura Institute of Materials Structure Science, High Energy Accelerator Research Organization, Tsukuba, Ibaraki 305-0801, Japan

[email protected]

Abstract: X-ray multiple image radiography is currently under development at the Photon Factory. The preliminary experiments of x-ray multiple image radiography were carried out at the vertical wiggler beamline BL-14B. Both absorption-contrast and phase-contrast tomograms of various specimens were successfully obtained at the wavelength of 0.112 nm.

1. Introduction X-ray imaging is a powerful method for visualizing the inner structures of various specimens non-destructively. At the Photon Factory (PF), several cutting-edge x-ray imaging techniques, such as interferometer-based imaging [1] and analyzed-based imaging [2-4], were successfully developed and applied for observing industrial materials and biological specimens. In general, although the sensitivity of the analyze-based imaging is usually smaller than that of the interferometer-based imaging, the analyzer-based imaging can cover a wider variety of samples than the interferometer-based imaging. Further, the analyzer-based imaging can produce both absorption-contrast and phase-contrast images from a set of images recorded at different angles of the analyzer crystal. Recently, it was shown that the analyzer-based imaging can also produce ultra-small angle x-ray scattering (USAXS) images reflecting microstructures in a sample [5]. This new x-ray imaging technique is called multiple image radiography, and is currently under development at the Photon Factory. In this paper, we show recent results obtained by the x-ray multiple image radiography. 2. Principle of x-ray multiple image radiography The x-ray optics of multiple image radiography is essentially same with that of the x-ray analyzer-based imaging as schematically shown in Fig. 1. This optics is composed of collimator crystal and analyzer crystal in non-dispersive (+, -) arrangement. The monochromatic x-rays are, at first, expanded by the collimator and then incident upon a sample on a rotation stage. The transmitted x-rays through the sample are analyzed by the analyzer, and then recorded by an x-ray area detector. In the x-ray multiple image radiography, it is necessary to scan the analyzer crystal over the Bragg

diffraction region, and to record diffracted image at each analyzer angle. The scanning angle of the analyzer is usually set several times larger than the angular region of the Bragg diffraction. The absorption-contrast, refraction-contrast (differential-phase-contrast) and phase-contrast images are obtained by the following equations:

( ) ( )∑ Δ∝j jabs yxIyxI θ,,, ,

( ) ( )∑∑ ΔΔ×Δ∝j jj jjrefract yxIyxIyxI θθθ ,,,,),(

Fig.1. X-ray optics of multiple image radiography


79

'),'(),(0

dxyxIyxIx

refractphase ∫∝

where Δθj is the offset angle of the analyzer from the Bragg angle and I(x, y, Δθj) is the recorded image at Δθj. From a set of the recorded images, we can also draw a rocking curve for each pixel at (x, y), and obtain the full width at half maximum (FWHM). Due to the effect of the USAXS in the sample, the FWHM observed at each pixel is slightly larger than that obtained without the sample. By mapping this increase of the FWHM, Δfwhm(x, y), we can obtain a USAXS-contrast image reflecting the microstructures in the sample. 3. Experimental and results Preliminary experiments of the x-ray multiple image radiography were performed at the vertical-wiggler beamline BL-14B of the Photon Factory. The white beam from the light source was monochromated at 0.112 nm by the Si(111) double-crystal monochromator. The incident monochromatic x-rays were expanded in the horizontal direction by an asymmetrically-cut Si(220) collimator crystal (α = 8°). As a sample, we observed a plastic tube filled with fine fibers. The diameter of the tube was about 3.8 mm. The transmitted x-rays through the sample were analyzed by a Si(220) crystal (α = 0°). The diffracted x-rays by the analyzer were recorded by an x-ray CCD camera (Photonic Science Ltd., XFDI) that consisted of a GdO2S:Tb scintillator, glass fiber plate, and CCD. The effective pixel size was 23 μm (H) × 23 μm (W) and the number of pixels were 1384 (H) × 1032 (V).

The analyzer was scanned from Δθ = -10 arcsec to Δθ = +10 arcsec in 1.0 arcsec steps. At each angle, the sample was rotated around the vertical axis from 0° to 180° in steps of 0.72°. The exposure time for each image was 150 msec. The total measurement time was about 2.5 hours. Figure 2 shows (a) absorption-contrast and (b) phase-contrast tomograms reconstructed by the filtered back-projection method. The phase-contrast tomogram of the plastic tube is much clearer than the absorption-contrast tomogram. On the other hand, the absorption-contrast tomogram of the fine fibers is almost as clear as the phase-contrast tomogram. This contrast enhancement is considered to be due to Fresnel diffraction and USAXS. The USAXS-contrast image is currently under analysis and will be reported elsewhere.

4. Conclusions The preliminary experiments of the x-ray multiple image radiography were carried out at the vertical wiggler beamline BL-14B of the Photon Factory. At the wavelength of 0.112 nm, both absorption-contrast and phase-contrast tomograms of the plastic tube filled with the fine fibers were successfully reconstructed by the filtered back-projection method. The phase-contrast tomogram of the plastic tube was much clearer than the absorption-contrast tomogram, whereas the absorption-contrast tomogram of the fine fibers is almost as clear as the phase-contrast tomogram. This contrast enhancement of the fine fibers is considered to be due to Fresnel diffraction and USAXS. Further analyses are currently under way and the results will be reported elsewhere. Acknowledgment This work was performed under the approval of the Program Advisory Committee of the Photon Factory (2013G054).

Fig. 2. (a) Absorption-contrast and (b) phase-contrast tomograms of the plastic tube filled with fine fibers. For the reconstruction, the filtered back-projection method was used. The diameter of the tube was about 3.8 mm.


80

References [1] A. Momose, T. Takeda, Y. Itai and K. Hirano, “Phase-contrast x-ray computed tomography for observing

biological soft tissues”, Nature Medicine 2(4), 473-475 (1996). [2] I. Koyama, Y. Hamaishi and A. Momose, “Phase tomography using diffraction-enhanced imaging”, AIP Conf. Proc.

705, 1283-1286 (2004). [3] A. Maksimenko, M. Ando, H. Sugiyama and T. Yuasa, “Computed tomographic reconstruction based on x-ray

refraction contrast”, Appl. Phys. Lett. 86, 124105 (2005). [4] K. Hirano, “X-ray angle-resolved computed tomography using an asymmetric analyzer crystal”, Jpn. J. Appl. Phys.

50, 026402 (2011). [5] E. Pagot, P. Cloetens, S. Fielder, A. Bravin, P. Coan, J. Baruchel, J. Hartwig and W. Thomlinson, “A method to

extract quantitative information in analyzer-based x-ray phase contrast imaging”, Appl. Phys. Lett., 3421-3423 (2003).


81

Development of nanosecond time-resolved Dispersive XAFS system for irreversible phenomena

Yasuhiro Niwa1,*), Tokushi Sato2), Kei Takahashi1), Masao Kimura1,3), Masahiko Hiraki4), Kohei Ichiyanagi1) 1) Photon Factory, Institute of Materials Structure Science, High Energy Accelerator Research Organization

2) Center for Free-Electron Laser Science, DESY, 3) SOKENDAI, 4)Mechanical Engineering Center, Applied Research Laboratory, High Energy Accelerator Research Organization

*[email protected]

Abstract: The nanosecond time-resolved dispersive XAFS (DXAFS) system combined with Nd:YAG laser has been developed to elucidate irreversible processes such as diffusions, dislocations, etc., at NW2A beam line in PF-AR in KEK, Japan. The fragmentation process of copper foil induced by high-power laser shock was observed using this system. The energy shift of copper K-edge in XANES spectra was not observed. The amplitude of EXAFS oscillation of the copper was gradually decreased up to 100 ns, and almost disappeared at 100 ns.

1. Introduction It is known that materials show unique structures under extreme conditions such as high pressure and high temperature. Using high-power laser as a trigger leads us to investigate the dynamics of irreversible structure changes in the extreme conditions, which is indispensable to understand many phenomena such as phase transition, fragmentation, and spin. Understanding the mechanism of the destruction of materials is very important from the point of view of the design of new materials and the degradations control of materials. A crack during the destruction process arises from microscopic changes such as elastic and plastic deformations in materials. The elastic deformation is reversible and detectable by XRD. On the other hand, the plastic deformation is irreversible and undetectable by XRD because it arises locally. X-ray absorption fine structure (XAFS) is a very powerful tool to obtain such local structural change. Dispersive XAFS (DXAFS) is one of special technique of XAFS specialized for dynamical observation; a XAFS spectrum of whole energy range of interest can be obtained using bent crystal (polychrometor) at once without any mechanical movements[1]. We have developed the single shot DXAFS system combined with the pulsed-laser. Since this system makes it possible to obtain XAFS spectra with only one pulse X-ray (single shot) from the synchrotron radiation source with sub-nano second time resolution in principle, it is possible to observe such an irreversible and locally-generated phenomenon. It is expected that early process of the destruction of materials such as cracking and deformation is revealed using our single shot DXAFS system. We carried out the DXAFS study of the laser shock-induced fragmentation of copper foil.

2. Experiments DXAFS measurements were carried out at NW2A beamline of PF-AR at KEK. NW2A is one of the most high flux hard X-ray beamline in PF and PF-AR because of its tapered undulator. The configuration of the system is shown in Fig. 1. A Si(111) with bending radius of 2 m was used as polychromator. A silicon microstrip (XSTRIP) was used as position sensitive detector of DXAFS. The laser pulse from a Nd:YAG laser (Powerlite-8000, Continuum, USA) with the pulse duration of 10 ns and the power of 0.2 TW cm-2 was used to induce a shock wave. The wavelength of the laser was 1064 nm. The timing among an X-ray pulse, laser pulse and spectrum detection was synchronized with the RF master clock through the delay generator (DG645). The DG 645 also controlled the delay timing between X-ray and laser pulse. The diameter of the laser and the X-ray at the sample position were 300 µm and 150 µm, respectively. The copper foil of 5 µm thickness was used as a sample. Since the copper was destructed and flies apart by laser irradiation, the sample was exchange every laser shot and measurement using our original sample changer. The XAFS spectrum corresponding delay time was obtained by integrating each spectrum.


82

Fig. 1 Schematic diagram of the new DXAFS system.

3. Results and Discussions The copper foil was laser-shocked with Nd:YAG laser, and time evolution of XAFS spectra and EXAFS oscillations around Cu K-edge was clearly obtained by changing delay times from 0 to 300 ns. Fig. 2 shows normalized XAFS spectra of various delay time. A spectrum obtained with conventional step-scan measurement was also shown as a reference.

Fig. 2 XAFS spectra of the Cu foil in various delay time.

The energy shift of absorption edge cannot be observed in measuring time delay. This indicates that the copper

keeps metallic state in spite of laser irradiation. The distinctive spectral structure of metallic copper is clearly obtained within 3 ns after laser irradiation. On the other hand, it was gradually decrease at 30 ns and almost disappeared at 300 ns. However XAFS oscillation is not appeared in case of gas phase sample in general and the spectrum is similar to that of 300 ns, the height of edge jump of the raw XAFS spectrum at 300 ns is more than 90% that of before irradiation. This means that the density of the copper after 300 ns of laser irradiation is almost comparable to solid state of copper. It is assume that the copper after 300 ns laser irradiation is not gas phase but solid state although the amplitude of oscillation of XAFS spectrum is very small. Fig. 3 shows radial structure functions (RDF). The peaks at 0.2 nm and around 0.4 nm in RDFs are interactions of Cu-Cu of nearest and second or more neighbors, respectively. However the peak at 0.2 nm is still remained, the peak around 0.4 nm is disappeared in 30 ns of RDF. It is suggested that the copper becomes smaller particle by the laser-shock. In the RDF of 300 ns, there are no peaks in all range because the EXAFS oscillation is almost disappeared.


83

Fig. 3 Radial structure function in various delay time.

4. Conclusions The laser shock experiment as an irreversible phenomenon of copper foil was observed using newly development single shot DXAFS system. The valence of the copper foil does not change and the copper keeps metallic state in several hundred nanosecond after laser irradiation. However the copper still keeps bulk state within 3 ns after laser irradiation, it fragments and becomes smaller particle after 30 ns. The cause for disappearance of EXAFS oscillation at 300 ns requires further investigation.

Acknowledgement A part of this research done by YN was carried out with the support of SIP (Cross-ministerial Strategic Innovation Promotion Program).

References [1] T. Matsushita and R. P. Phizackerley, Jpn, J. Appl. Phys., 20, 2223 (1981).


84

X-ray computed tomography imaging of structure materials using synchrotron radiation

Yumiko Takahashi, Keiichi Hirano, Kazuyuki Hyodo and Masao Kimura Institute of Materials Structure Science, High Energy Accelerator Research Organization, Tsukuba, Ibaraki 305-0801, Japan.

Phone: +81-29-879-6290, Fax: +81-29-864-2801 (ex. 6840) E-mail: [email protected]

Abstract: Synchrotron radiation X-ray computed tomography was applied to visualize the internal structure of SiC fiber-reinforced SiC matrix (SiC/SiC) composite. In addition to the conventional absorption contrast image, phase contrast images were clearly observed. These images revealed the fiber bundle structure and the pore distribution in the SiC/SiC composite.

1. Introduction Silicon carbide fiber-reinforced silicon carbide matrix (SiC/SiC) composite has attracted a growing interest for the structural engineering material, due to its excellent thermal and chemical stability [1]. Since non-destructive characterization of the SiC/SiC is important to the characteristic elucidation and improvement, the 3D geometrical characterization of the SiC/SiC by synchrotron radiation x-ray computed tomography (CT) was performed in order to verify the possibilities of this technique.

When compared to conventional laboratory x-ray source, synchrotron radiation has several fundamental advantages for CT measurements. One of the important one is that, monochromatic x-rays can be obtained by using a single crystal called monochromator, which is used to select the x-ray energy. By the selection of x-ray energy, the transmittance can be optimized for the examination of each sample. Monochromaticity also reduces the beam hardening effect that is usually observed in laboratory x-ray source. This effect causes artifacts in the reconstructed images, and deteriorates the quality of quantitative characterization. As other merits of synchrotron radiation, high intensity allows for short scan times and strong collimation of the beam contributes to high spatial resolution of images [2]. Moreover, in addition to the conventional CT (absorption contrast CT), the phase contrast CT technique highlighting small differences of refractive index within structures is suitable for the synchrotron radiation source [3].

In this study, both absorption contrast CT and phase contrast CT were applied to the SiC/SiC.

2. Experiment The experiments were performed at the vertical wiggler beamline BL-14B of the Photon Factory, High Energy Accelerator Research Organization (KEK). The x-ray energy was adjusted to 20.6 keV using a Si(111) double-crystal monochromator. The incident beam was collimated and expanded in the horizontal direction by an asymmetrically cut Si(220) crystal (α = 8 deg) to produce an x-ray beam with uniform intensity. The x-rays transmitted through a sample were expanded by a Si(220) asymmetric analyzer crystal (α = 6 deg, magnification factor is about 5) in the horizontal plane, then the x-rays diffracted by the analyzer were recorded on an x-ray charge-coupled device (CCD) camera (Photonic Science, VHR). The x-ray CCD camera consisted of a GdO2S:Tb scintillator, a glass fiber plate (taper =1:1) and a CCD sensor. The effective pixel size was 7.4 μm (H) ⋅ 7.4 μm (V), and the number of pixels were 4872 (H) ⋅ 3248 (V). A series of radiographic images taken with rotating the sample around the vertical axis are reconstructed via filtered back-projection (FBP) method.

In the case of the phase contrast CT, diffraction enhanced x-ray imaging (DEI) method was employed [4-6]. In DEI-CT two image sets were acquired on the slopes of the analyzer rocking curve at the angular positions θB ± (ΔθD)⁄2, where θB is the Bragg angle and ΔθD is the full-width at half-maximum of the rocking curve. From these two image sets the phase images were obtained by Hirano’s method before the FBP [7].

The SiC/SiC in the shape of 2.5 ⋅ 2.8 ⋅ 9.6 mm3 was prepared for the sample.

3. Results and discussion Volume rendering of the SiC/SiC sample is shown in Fig. 1(a). The fiber bundle structure is visible. It is possible to observe arbitrary cross section by virtual cut of image to detect matrix defects such as pores and cracks in the volume and to follow the fiber undulations. Figure 1(b) shows the porous network that was extracted from fig. 1(a). The individual fibers were, however, not detectable due to the resolution limit of about 15 μm in this image, which was mainly determined by the resolution of detector.

Figure 2 shows two-dimensional CT images for the directions of perpendicular to the z-axis in Fig. 1. Figure 2(a) is an absorption contrast CT, where the contrast between the sample constituents is due to the intrinsic differences in linear attenuation coefficients. Figure 2(b) reconstructed from the DEI-CT data shows the map of phase-shift that occur


85

when x-rays are refracted from their initial path while traversing the sample. For the materials consisting of light elements, the DEI generally provides improved contrast compared to the absorption radiography. However, the fiber bundle structure of Fig. 2(a) was unexpectedly clearer than Fig. 2(b). Since a phase map is considered as a projection of the density distribution of this sample, Fig. 2(b) indicates the homogeneity in the fiber bundle. Consequently, the edge effect may play the important role to generate the contrast in Fig. 2(a).

4. Conclusion The internal structure of SiC/SiC composite was observed by the absorption contrast CT and the DEI-CT using synchrotron radiation. The results showed the further optimization of measuring conditions, especially in DEI-CT measurements will give us unique information to visualize the internal structure of the SiC/SiC in high quality, which is difficult to obtain by conventional laboratory-type CT apparatus. More quantitative analysis also will be carried out for the next step.

Fig. 1. (a) 3D volume rendering image and (b) porous network of the SiC/SiC. Scale bars indicate 1mm.

z

x

y

Fig. 2. (a) Absorption contrast CT and (b) phase map obtained by DEI-CT of the SiC/SiC.

1 mm

(a) (b)

(a) (b)


86

Acknowledgment

This work was performed under the approval of the Program Advisory Committee of the Photon Factory (2015S2-002).

References

[1] R. Naslain, “Design, Preparation and Properties of Non-Oxide CMCs for Application in Engines and Nuclear Reactors: An Overview”, Compos. Sci. Technol., 64 (2), 155–170 (2004).

[2] S. Mobilio, F. Boscherini, C. Meneghini, (Eds.) [Synchrotron Radiation Basics, Methods and Applications], Springer-Verlag Berlin Heidelberg, (2014). pp 594-595.

[3] R. Fitzgerald, Phase Sensitive X Ray Imaging , Phys. Today, 53(7), 23-26 (2000).

[4] K. Hirano, Y. Takahashi and H. Sugiyama, “Application of variable-magnification X-ray Bragg magnifier to analyzer-based phase-contrast computed tomography”, Jpn. J. Appl. Phys. 53, 040302 (2014).

[5] D. Chapman, W. Thomlinson, R. E. Johnston, D. Washburn, E. Pisano, N. Gmür, Z. Zhong, R. Menk, F. Arfelli and D. Sayers, “Diffraction enhanced x-ray imaging”, Phys. Med. Biol., 42, 2015-2025 (1997).

[6] F. A. Dilmanian, Z. Zhong, B. Ren, X. Y. Wu, L. D. Chapman, I. Orion and W. C. Thomlinson, “Computed tomography of x-ray index of refraction using the diffraction enhanced imaging method”, Phys. Med. Biol., 45, 933-946 (2000).

[7] K. Hirano, Month. J. Med. Imaging Inf., 38, 1271-1276 (2006) (in Japanese).


87

Full-field deformation measurement of carbon fiber reinforced plastic under three-point bending test at micron scale

Qinghua Wang 1), Shien Ri 1), Takashi Tokizaki 1), Yoshihisa Tanaka 2) 1) Research Institute for Measurement and Analytical Instrumentation, National Institute of Advanced Industrial Science

and Technology, Tsukuba, Ibaraki 305-8568, Japan. Tel: 81-29-861-3631, email: [email protected] 2) Hybrid Materials Unit, National Institute for Materials Science, Tsukuba, Ibaraki 305-0047, Japan

Abstract: Full-field deformation distributions at the micron scale of a carbon fiber reinforced plastic (CFRP) specimen under three-point bending were investigated from the scanning moiré method and theoretical analysis. The scanning moiré fringes were generated from the interference between the micron specimen grating and the scanning lines in a scanning electron microscope. The displacements relative to the scanning lines in the x and the y directions were obtained from the scanning moiré fringes. The normal strains in the x and the y directions as well as the shear strain were measured from the scanning moiré method. The maximum and the minimum principle strains as well as their orientations were determined theoretically from the strain status analysis. The combination of the scanning moiré method and the theoretical analysis provides a good way to explore the full-field principle strain distributions and their orientations, regardless of the specimen grating direction. The proposed combination method is useful to understand the strain status of materials in the full field quantitatively and non-destructively. 1. Introduction Recently, carbon fiber reinforced plastics (CFRP) have drawn great attention in both science and industries owing to the advantages of high strength and lightweight [1]. They become increasingly popular in a variety of fields such as automobiles, motorcycles, aerospace and computers. With the development of miniaturization, the feature dimensions of CFRPs decrease to the micron or even nanoscale. To evaluate their mechanical properties and failure mechanisms such as crack initiation and propagation, it is indispensable to measure their full-field deformation distributions at the micron/nanoscale.

The commonly used optical techniques for full-field deformation measurement at the micron/nanoscale include the moiré methods [2-4], the geometric phase analysis [5], the digital image correlation method [6], and the speckle interferometry [7]. The geometric phase analysis is limited by the small view field with size of only 10~100 times the grating pitch, because the specimen grating has to be observed as the deformation carrier. The last two techniques have a large view field, however, the measurement repeatability is low as the speckle is easily influenced by the noise in the digital image correlation method, and the measurement is extremely weak to vibration in the speckle interferometry. Among the moiré methods, the moiré interferometry [8] needs complicated optical paths and good anti-vibration measures, the sampling moiré method [9] has disadvantage of a small view field, and the digital moiré method [10] is only suitable for strain greater than 0.01. Fortunately, the microscope scanning moiré method possesses a large view field with size of usually 500~2000 times the specimen grating pitch, high measurement repeatability, high accuracy and simple operation [11]. In this study, we will use the scanning electron microscope (SEM) scanning moiré method [12-14] to measure the normal and the shear strain distributions of a CFRP specimen under three-point bending test.

For more detailed understanding of the strain status of CFRP [15], we will calculate the maximum and the minimum principal strains and their orientations based on theoretical strain status analysis and the strain distributions measured by the SEM scanning moiré method. The combination of the scanning moiré method and the theoretical analysis is robust in evaluating the full-field strain status of materials, no matter whether the specimen grating direction is parallel to the axial direction of the tested object or not.

2. Principles of methods

2.1 Scanning moiré method The scanning moiré fringes are caused by the interference between a specimen grating and the microscope scanning lines acting as the reference grating (Fig. 1). The displacement of the specimen relative to the scanning lines [12] can be obtained from

urela=mT (1)

where m is an integer (m=…,-1,0,1,2,…) indicating the moiré fringe order, and T is the scanning spacing. For two-dimensional deformation measurement, the specimen or the scanning lines should be rotated by 90°. From the moiré fringes before and after rotation, both the relative displacements in the x and the y directions can be measured.

The normal strains and the shear strain of the specimen relative to the scanning lines are obtainable from the first-order differentials of the relative displacements

_ rela _ rela_ rela _ rela= , =x y

x y

u u

x yε ε

∂ ∂

∂ ∂ (2)


88

_ rela _ rela_ rela = x y

xy

u u

y xγ

∂ ∂+

∂ ∂ (3)

The normal strains relative to the scanning lines and the actual normal strains of the specimen can also be calculated from the relationship between the specimen grating pitch before and after deformation and the scanning spacing. Taking the normal strain in the y direction as an example, the normal strain relative to the scanning lines and the actual normal strain of the specimen are presented in Eq. (4) and Eq. (5), respectively.

'

_ relay

y

p T

Tε

−= (4)

'y y

yy

p p

pε

−= (5)

where py and py’ are the specimen grating pitch before and after deformation, respectively (Fig. 2). From Eqs. (4) and (5), the actual normal strain of the specimen in the y direction can also be expressed as

_ rela(1 ) 1y yy

T

pε ε= + − (6)

Similarly, the actual normal strain of the specimen in the x direction is measureable from the following equation

_ rela(1 ) 1x xx

T

pε ε= + − (7)

Because the angle variation of the specimen grating relative to the scanning lines and the actual angle variation are the same (Fig. 2), we know that the actual shear strain of the specimen is equal to the relative shear strain.

_ rela=xy xyγ γ (8)

Fig. 1. Formation principle of the microscope scanning moiré method

Fig. 2. Schematic diagram of the geometrical relationship and the strain status including principle and shear strains

2.2 Strain status analysis For the plane stress problems, the normal strains (linear strains) and the shear strain exist simultaneously. The principle strains are more useful to understand the strain status and the potential fracture area. The maximum and the minimum principle strains [16] can be determined by

22

max ( )2 2 4

x y x y xyε ε ε ε γε

+ −= + + (9)

22

min ( )2 2 4

x y x y xyε ε ε ε γε

+ −= − + (10)

The orientations of the maximum and the minimum principles strains are perpendicular to each other. The orientation of the maximum principle strain can be calculated from

0tan 2 xy

x y

γα

ε ε=

− 11)

The principle strains and their orientations are independent of the directions of the normal strains.


89

3. Materials and experiments

3.1 Specimen preparation and grating fabrication The CFRP specimen was made up of IM600 and K13D carbon fibers mixed with epoxy resins. All the fibers were parallel, and the IM600 layer and the K13D layer were arranged alternately as shown in Fig. 3(a). The diameter of the IM600 carbon fiber was 5 μm, and the thickness of the IM600 layer was approximately 110 μm. The diameter of the K13D carbon fiber was within 10~11 μm, and the thickness of the K13D layer was approximately 300 μm.

The specimen was prepared with mechanical cutting using an abrasive wheel cutting machine. The thickness and the width of the specimen were t=1 mm and b=5 mm, respectively. The length of the sample was approximately 30 mm. The 1mm×30mm surface of the specimen was polished using an automatic polishing machine with sandpapers and polishing solutions successively. The fiber length direction was perpendicular to the polished surface.

After polishing, a cross grating with pitch of p=3.72 μm was fabricated by UV nanoimprint lithography (EUN-4200 device). The grating fabrication process is displayed in Fig. 4. The used resist was PAK01, and the UV illustration time was 30s. The grating area was around 1mm×20mm. After grating fabrication, the specimen surface was coated with gold with thickness of less than 30 nm for conductivity. The appearance of the fabricated grating is exhibited in Fig. 5.

3.2 Three-point bending test in SEM The three-point bending test was performed using a simple loading jig which could be placed in the used SEM (FEI Quanta 200FEG). The support span was L=16 mm, 16 times the thickness (1mm), according to the standards of American Society for Testing and Materials (ASTM, No. D 790-03). A strain gauge was affixed on the 5mm×30mm surface, below the loading point, to measure the maximum tensile strain of the specimen, as seen in Fig. 3(a). When the strain gauge value was 0.008, the specimen with grating was observed in the SEM for full-field deformation measurement. The structure of the three-point bending test in the SEM is presented in Fig. 3(b).

Fig. 3. (a) Sketch map and (b) SEM image of the three-point bending loading structure of the CFRP specimen

Fig. 4. Fabrication process of a micron grating using UV nanoimprint lithography

Fig. 5. SEM image of the micron cross grating with pitch of 3.72 μm on CFRP

Fig. 6. SEM scanning moiré fringes on CFRP when the scanning spacing is 3.60 μm, and the scanning direction is

parallel to the axial direction under three-point bending


90

To get distinct SEM scanning moiré fringes for deformation measurement, the moiré formation condition should be met, i.e., the scanning spacing is within 0.8~1.2 times the specimen grating pitch, and the angle between the scanning lines and the specimen grating lines is small or zero. From our experiments, when the magnification of the SEM is 140×, the scanning spacing is T=3.60 μm, 0.97 times the specimen grating pitch. Under this circumstance, the SEM scanning moiré fringes emerged when the scanning direction is parallel to the axial direction of the CFRP specimen in the area just below the loading point (Fig. 6). Although the moiré fringes are clear, the fringes are too dense in Fig. 6 which will bring great error in deformation measurement.

The reason why the moiré fringes in Fig. 6 are dense lies in that the specimen grating lines are not parallel to the axial direction of the specimen due to the positioning error during grating fabrication. In this case, we can rotate the specimen or the scanning direction to find a better moiré pattern, and then use the methods explained in Section 2 to calculate the principle strains for understanding the strain status.

4. Results and discussion

4.1 Moiré observation and relative displacement distributions For improving the deformation measurement accuracy and visually understanding the characteristics of the deformation distribution, we rotated the SEM scanning lines by θ=4.5°clockwisely to make the scanning lines be almost parallel to one of the grating directions. For two-dimensional deformation measurement, the scanning lines should be continuously rotated by 90° to get the scanning moiré fringes in another direction. The ux-field SEM scanning moiré fringes when the scanning direction (y direction) deviates 94.5° from the axial direction and uy-field moiré fringes when the scanning direction (x direction) deviates 4.5° from the axial direction are illustrated in Figs. 7(a) and 7(b), respectively. It can be seen that, the moiré fringes are distinct and visually reflect the displacement distributions of the specimen grating relative to the scanning lines, because the moiré fringes indicate the displacement contours.

(a) ux-field moiré fringes when the scanning direction deviates 94.5° from the axial direction

(b) uy-field moiré fringes when the scanning direction deviates 4.5° from the axial direction

Fig. 7. SEM scanning moiré fringes on CFRP when the scanning spacing is 3.60 μm and the scanning direction is approximately parallel to the grating direction under three-point bending

Fig. 8. Relative displacement distributions of CFRP relative to the scanning lines under three-point bending (1548μm×612μm)


91

The two moiré patterns in Fig. 7 were used to measure the displacement fields by the SEM scanning moiré method. A same region with size of 1548μm×612μm labeled by the black rectangle in Fig. 7 was chosen as the region of interest (ROI), avoiding the influence of the contaminated spot in the bottom area. Using Eq. (1), the relative displacement distributions in the x and the y directions in the ROI of the CFRP specimen under three-point bending were measured, as seen in Fig. 8. The relative displacements are useful to evaluate the normal and the shear strains of the specimen.

4.2 Distributions of normal and shear strains Based on Eqs. (2) and (3), the relative normal and shear strains of the specimen relative to the scanning lines could be calculated from the first-order differentials of the relative displacements in Fig. 8. Furthermore, the actual normal and shear strains of the CFRP specimen were measurable using Eqs. (6)-(8). The measured actual normal strain distributions in the x and the y directions and the actual shear strain distribution are plotted in Fig. 9.

The normal strain in the x direction is compressive in the upper area and tensile in the lower area, and the normal strain in the y direction is compressive in the whole area from Fig. 9. This phenomenon basically agrees with the typical strain characteristics under three-point bending. However, the CFRP specimen has its individual deformation features as a composite material. The normal strain in the x direction is maximal in the lower-left area and decreases gradually along the upper-right and upper-left directions. The normal strain in the y direction is minimal in the lower-left (close to lower-middle) area and increases gradually along the upper-right (close to rightwards) and upper-left directions. The shear strain is positive and its distribution characteristic is similar to that of the normal strain in the x direction.

4.3 Distributions of principle strains and orientations The maximal and the minimal principle strains and their orientations are vital parameters for evaluating deformation characteristics, potential crack initiation and propagation, and instabilities. Based on the actual normal and shear strains in Fig. 9, the principle strains and their orientations were determined using Eqs. (9)-(11). The full-field distributions of the maximal and the minimal principle strains are presented in Fig. 10.

The maximum principle strain in Fig. 10(a) is tensile and has almost the same distribution characteristic as the shear strain in Fig. 9(c) except the values. It indicates that the shear strain contributes most to the maximum principle strain. The minimum principle strain in Fig. 10(b) is compressive and has greater absolute values than the maximum principle strain, demonstrating that the minimum principle strain is more likely to induce instability, and the most unstable region is at the upper-right corner. The maximum and the minimum principle strains with orientations (Fig. 11) facilitate us to understand the strain distribution features deeply. It should be noted that, when the scanning direction is rotated, only the normal and the shear strains will vary, and the principle strains as well as their orientations will not be affected by the rotation angle of the specimen or the scanning lines.

Fig. 9. Distributions of the normal strains and the shear strain of CFRP under three-point bending (1548μm×612μm)

Fig. 10. Principle (maximum and minimum) strain distributions of CFRP under three-point bending (1548μm×612μm)


92

Fig. 11. Principle strain distributions with orientations of CFRP under three-point bending (1548μm×612μm)

5. Conclusions The deformation distributions including the normal, shear and principle strains of CFRP under three-point bending were investigated by combination of the SEM scanning moiré method and the theoretical strain status analysis. The normal strain in the x direction, the shear strain and the maximum principle strain have similar distribution characteristics except the values. The minimum principle strain is more likely to trigger instability, and the most unstable area is at the upper-right corner under three-point bending. The measurement method presented in this study provides an effective and robust approach for micron-scale and full-field deformation evaluation of a variety of materials even if the specimen grating is along an arbitrary direction.

Acknowledgement The authors acknowledge the financial support from Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office. The authors are also grateful to Dr. Satoshi KISHIMOTO in NIMS for offering the nanoimprint mold and Dr. Kimiyoshi NAITO in NIMS for providing the CFRP specimen.

References [1] K. Naito, J.M. Yang and Y. Kagawa, “Tensile properties of high strength polyacrylonitrile (PAN)-based and high modulus

pitch-based hybrid carbon fibers-reinforced epoxy matrix composite”, J. Mater. Sci., 47(6), 2743-2751 (2012).

[2] S. Kishimoto, M. Egashira and N. Shinya, "Microcreep deformation measurements by a moiré method using electron beam lithography and electron beam scan", Opt. Eng. 32(3), 522-526 (1993).

[3] Q.H. Wang, H. Tsuda and H.M. Xie, “Developments and applications of moire techniques for deformation measurement, structure characterization”, Recent Pat. Mater. Sci., 8(3), 1-20 (2015).

[4] Y. Kondo and E. Okunishi, “Magnified pseudo-elemental map of atomic column obtained by Moiré method in scanning transmission electron microscopy”, Microscopy, 63(5), 391-395 (2014).

[5] Q.H. Wang, S. Kishimoto, H.M. Xie, Z.W. Liu and X.H. Lou, "In situ high temperature creep deformation of micro-structure with metal film wire on flexible membrane using geometric phase analysis", Microelectron. Reliab., 53(4), 652-657 (2013).

[6] Z.X. Hu, H.M. Xie, J. Lu, T. Hua and J.G. Zhu, "Study of the Performance of Different Subpixel Image Correlation Methods in 3d Digital Image Correlation", Appl. Opt., 49(21), 4044-4051 (2010).

[7] R. Erf, (Ed.). [Speckle metrology], Elsevier, (2012).

[8] A. Post., B. Han, P. Ifju, [High sensitivity moiré: experimental analysis for mechanics and materials], Springer Science & Business Media, (2012).

[9] S. Ri, T. Muramatsu, M. Saka, K. Nanbara and D. Kobayashi, "Accuracy of the Sampling Moiré Method and Its Application to Deflection Measurements of Large-Scale Structures" Exp. Mech., 52(4), 331-340 (2012).

[10] Q.H. Wang, S. Kishimoto and Y. Yamauchi, “Three-directional structural characterization of hexagonal packed nanoparticles by hexagonal digital Moiré method”, Opt. Lett., 37(4), 548-550 (2012).

[11] H.M. Xie, Q.H. Wang, S. Kishimoto and F.L. Dai, “Characterization of planar periodic structure using inverse laser scanning confocal microscopy moiré method and its application in the structure of butterfly wing”, J. Appl. Phys., 101(10), 103511 (2007).

[12] Q.H. Wang and S. Kishimoto, "Simultaneous analysis of residual stress and stress intensity factor in a resist after UV-nanoimprint lithography based on electron moiré fringes," J Micromech. Microeng., 22(10), 105021 (2012).

[13] S. Kishimoto, “Electron moiré method”, Theor. Appl. Mech. Lett., 2(1), 011001 (2012).

[14] Q.H. Wang, S.Kishimoto, X.F. Jiang Yamauchi, Y., "Formation of secondary moiré patterns for characterization of nanoporous alumina structures in multiple domains with different orientations", Nanoscale, 5(6), 2285-2289 (2013).

[15] Q.H. Wang, S. Ri, H. Tsuda, S. Kishimoto, Y. Tanaka and Y. Kagawa, “Deformation measurement of carbon fiber reinforced plastics using phase-shifting scanning electron microscope moiré method after Fourier transform”, Proc. icOPEN, 9524-43 (2015).

[16] R.C. Hibbeler, [Statics and Mechanics of Materials], Pearson Higher Ed., (2013)


93

Laser ultrasonic testing of carbon –fiber-reinforced plastic (CFRP) with Mid IR pulsed light source generated by OPO

Makoto Watanabe


94

Poster papers

IMASM Theme 2 Trace Light Elements


95

Superconducting X-ray spectrometer for trace light elements in structural materials

Shigetomo Shiki1), Go Fujii1), Masahiro Ukibe1), Yoshinori Kitajima2), Masataka Ohkubo1) 1) National Institute of Advanced Industrial Science and Technology, 1-1-1 Umezono Tsukuba, Ibaraki 305-8568, JAPAN

2) High Energy Accelerator Research Organization, JAPAN

Abstract: Investigating local atomic structure and chemical states (electronic structure) of trace light elements (B, C, N, O) in structural materials are important to improve mechanical performance. Fluorescent yield X-ray absorption fine structure (XAFS) spectroscopy is a method for analysing electron structure of specific elements in matrices. We developed an superconducting XAFS apparatus (SC-XAFS) with a 100-pixel superconducting tunnel junction (STJ) detector array with a high sensitivity and a high resolution for soft X-rays in order to analyze light-element dopants in materials. We report the properties of the SC-XAFS apparatus and first measurement result of X-ray absorption spectrum of nitrogen dopant in heat-resistant steels.

1. Introduction Heat-resistant steels are important structural materials for power stations and other social-infrastructures. The mechanical performance of the heat-resistant steels is controlled by adding some impurities. It is known that trace light elements such as boron and nitrogen drastically improve creep lifetime [1]. However, the understanding of the role of the light elements is not known well. The local structure and chemical states (electronic structure) of trace elements can be analysed by fluorescent yield X-ray absorption fine structure (XAFS) spectroscopy. However, in a soft X-ray region, it is difficult to obtain XAFS spectra of light elements of low density by using conventional semiconductor X-ray detectors, because the K-lines of light elements in matrices cannot be clearly resolved. In contrast, superconducting tunnel junction (STJ) detector is a promising device for the soft X-ray spectroscopy because of its superior characteristics: high energy resolving power, high counting rate capability, and high sensitivity [2-3]. A fluorescent yield XAFS spectroscopy apparatus using 100-pixel STJ detector (SC-XAFS) is constructed to investigate the nano-scale local structure of trace light elements in a soft X-ray region [4]. We report the current status of SC-XAFS instrumentation and the first results of X-ray absorption spectrum of the nitrogen dopant in a heat-resistant steel. 2. Experiment 2.1. Soft X-ray Spectrometer with Superconducting Tunnel Junction Detector The sample is placed in a UHV chamber with a base pressure of 10-7 Pa. The sample surface can be set to any angle between 0 and 90 degrees against synchrotron radiation X-ray beams. The STJ detector surface is set in parallel to the X-ray beams. The distance between the beam spot on the sample and the STJ detector is approximately 30 mm. The STJ detector is cooled using a cryogen-free 3He cryostat consisting of a dual stage pulse tube refrigerator and a 3He sorption cooler. The base temperature of the cryostat is 300 mK. The pixel size of the STJ detector is 100 x 100 μm2. Therefore the total sensitive area of the STJ detectors is 1 mm2. Each STJ pixel is connected to a readout circuit consisting of a charge sensitive amplifier and a pulse height analyser. The SC-XAFS is installed in soft-X-ray beam lines at the Photon Factory (PF), High-Energy Accelerator Research Organization (KEK). Available energy range of the synchrotron radiation is 70-5000 eV by selecting suitably beam line from BL-11A, BL-11B, or BL-16A in KEK-PF [4]. Currently, the SC-XAFS exhibits superior performances of an energy resolution of 7 eV in full width at half maximum in average over 100-pixel, and a counting rate of 500 k cps [5-9]. 2.2. Heat-resistant steel samples A heat-resistant steel sample contains 9% Cr, 150 ppm B, 300 ppm N in weight. The surface of the sample is mirror polished. XAFS spectra in this paper were obtained at BL-11A, which has a bending magnet source and a monochrometer for a range of 70-2000 eV. 3. Results and Discussions Fluorescent X-ray spectrum from the sample irradiated by X-ray of 450 eV is shown on Fig. 1(a). The N-K line (392 eV) is clearly separated from the C-K line (277 eV). The Fe-L lines (705 eV, 615 eV) and O-K line (525 eV) are generated by higher order X-ray from the monochrometer. The STJ X-ray spectrometer can measure a fluorescence yield of a specific light element while efficiently eliminating the influence of the fluorescent X-ray from other elements. Figure 1 (b) shows an XAFS spectrum of the 150 ppm nitrogen dopant in the sample. Unfortunately, it was impossible for us to evaluate the reliable chemical state of nitrogen using this data, because the energy resolution and photon flux of the monochrometer is not enough due to beamline troubles during the measurement time. However we believe that the quality of the XAFS spectrum will be improved, when the monochrometer in the beamline is recovered well, or it is


96

possible to use high performance beamlines. Acquisition of N-K edge XAFS spectra was successful at another beam line, and the latest result is reported by P. Fons et al. in these proceedings.

(a) (b) Fig. 1(a) Fluorescence X-ray spectrum of high temperature steel with nitrogen weight concentration of 150 ppm irradiated by 450 eV X-ray. (b) A fluorescent yield XAFS spectrum of a high temperature steel with nitrogen weight concentration of 150 ppm. 7. Conclusions STJ X-ray spectrometer is an indispensable device for material analysis of structural materials. The SC-XAFS works properly in the photon factory in KEK and exhibits superior performances of high energy resolution of 7 eV FWHM in average, high counting rate of 500 k cps, and a large detection area of 1 mm2. The SC-XFAS can measure absorption spectrum of nitrogen dopant in high temperature steel. In future, boron dopant will be measured after instrumental improvement. Acknowledgement This research was supported by Cross-ministerial Strategic Innovation Promotion Program (SIP). We thank Dr. F. Abe for sample preparations and fruitful discussions.

References [1] F. Abe, M. Tabuchi, S. Tsukamoto, “Mechanisms for boron effect on microstructure and creep strength of ferritic power plant

steels”, Energy Materials 4, 166-175 (2012)

[2] Friedrich, S., “Cryogenic X-ray detectors for synchrotron science”, J. Synch. Rad. 13, 159-171 (2006)

[3] P. Fons, H. Tampo, A. V. Kolobov, M. Ohkubo, S. Niki, and J. Tominaga, “Direct Observation of Nitrogen Location in Molecular Beam Epitaxy Grown Nitrogen-Doped ZnO”, Phys. Rev. Lett. 96, 045504 (2006)

[4] S. Shiki, N. Zen, M. Ukibe, and M. Ohkubo, “Soft X Ray Spectrometer Using 100 Pixel STJ Detectors for Synchrotron Radiation”, AIP Conf. Ser. 1185, 409-412 (2009)

[5] S. Shiki, M. Ukibe, N. Matsubayashi, N. Zen, M. Koike, Y. Kitajima, M. Ohkubo, “Current Status of AIST X-ray-Absorption-Spectroscopy (XAFS) Instrument with 100-Pixel Superconducting-Tunnel-Junction Array Detector”, J. Low Temp. Phys. 176, 604-609 (2014)

[6] M. Ukibe, S. Shiki, Y. Kitajima, and M. Ohkubo, “Soft X-ray detection performance of superconducting tunnel junction arrays with asymmetric tunnel junction layer structure”, Jpn. J. Appl. Phys. 51, 010115 (2012)

[7] S. Shiki, M. Ukibe, Y. Kitajima, and M. Ohkubo, “X-ray Detection Performance of 100-pixel Superconducting Tunnel Junction Array Detector in the Soft X-ray Region”, J. Low Temp. Phys. 167, 748–753(2012)

[8] M. Ohkubo, S. Shiki, M. Ukibe, N. Matsubayashi, Y. Kitajima, S. Nagamachi, “X-ray absorption near edge spectroscopy with a superconducting detector for nitrogen dopants in SiC”, Sci. Rep. 2, 831 (2012)

[9] G. Fujii et. al, in preparation


97

Energy-Dispersive X-Ray Detectors Based on Three-Dimensional Superconducting-Tunnel-Junctions

for Multi-Element Analysis of Trace Light Elements

Go Fujii1), Masahiro Ukibe1), Shigetomo Shiki1), and Masataka Ohkubo1) 1) National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki, 305-8568, Japan

TEL: +81-29-861-1393, FAX: +81-29-861-5088 E-mail: [email protected]

Abstract: Our energy-dispersive X-ray detectors using 100-pixel array of superconducting-tunnel-junction (STJ), of which a total detection area size is 1 mm2, have demonstrated excellent X-ray detection performance: na energy resolution of about 10 eV in full width at half-maximum (FWHM) and a counting rate of more than 100 k cps in a soft X-ray energy range, and is now used for X-ray absorption fine structure (XAFS) analyses. In order to achieve a high throughput analysis for impurities of dopants with a low concentration, the total detection area should be further increased up to several 10 times larger. We developed the STJ of three-dimensional structure (3D-STJ) in which the STJ pixel layer and the wiring lead layer is separated. The 3D-STJs can achieve a large detection area more than 10 mm2 by realizing more than 3000-pixels with close-packed arrangement. The 3D-STJ has a leak current of 40 nA and an energy resolution of 13 eV FWHM for the O-Kα line.

1. Introduction X-ray absorption fine structure (XAFS) analyses are used in synchrotron radiation facilities to investigate oxidation state, local coordination environment, spin state, electronic structure, and bond characteristics for specific elements in matrices. Particularly, the XAFS analyses in soft X-ray region is useful for investigating trace light elements in structural materials. However, in a soft X-ray region, the energy resolving power in general energy-dispersive X-ray detectors such as silicon drift detectors (SDDs)[1] and Si(Li) detectors is insufficient to achieve clear separation of the K-lines and L-lines of different elements.

Energy-dispersive X-ray detectors based on superconducting-tunnel-junctions (STJs) have demonstrated an excellent energy resolution, a high detection efficiency, and a high counting rate in a soft X-ray energy range[2],[3]. In fact, our X-ray detector using the 100-pixel array of Nb/Al/AlOx/Al/Nb STJs with a junction area of 100 µm square already has achieved a mean energy resolution of better than 10 eV in full width at half-maximum (FWHM), a relatively large detection area of 1 mm2, and a counting rate of more than 100 k cps. An XAFS instrument equipped with our STJ array detector (SC-XAFS) was installed in the synchrotron radiation beam line, and has been used for routine analyses. By utilizing SC-XAFS, it is possible to obtain XAFS spectra of trace light elements such as nitrogen dopants of 300 ppm in a SiC compound semiconductor [4]. That high performance is beginning to reveal the role of nitrogen and boron atoms in heat-resistant steels.

In order to achieve a high throughput analysis for trace light elements such as dopants in structural omaterials, the detection area of the STJ array detectors should be further increased up to several 10 mm2, which is several 10 times larger than the current one. We have introduced three-dimensional (3D) structure by embedding a wiring layer in a SiO2 layer underneath STJ pixels. In conventional layout, a junction layer and a wiring layer were arranged on same plane. In the design using conventional layout, the maximum pixel number is about 600 on 10 mm square chip, which is the maximum size for the STJ array chip mounting in our XAFS instrument. In contrast, by using the 3D structure, the maximum pixel number can be enlarged to more than 3000 [5]. In this work, we have made the 100-pixel array of the 3D-STJ and evaluated detection performance for soft X-rays.

2. Experiment

2.1 Fabrication The 3D-STJs were fabricated using conventional photolithographic techniques, DC magnetron sputtering, reactive ion etching (RIE), wet-etching process, caldera planarization[6], TEOS chemical vapor deposition, and lift-off process in the clean room for analog-digital superconductivity (CRAVITY) at National Institute of Advanced Industrial Science and Technology (AIST) [7]. The detail of the fabrication process of the 3D-STJs is described in the literature [8]. Fig. 1 shows an optical microscope image of fabricated 100-pixel STJ array. The layer structure of the junctions consisted of Nb (100 nm)/Al-AlOx (70 nm)/Al (70 nm)/Nb (300 nm). The critical current density of the STJs was designed about 200 A/cm2. The STJs with the size of 100 μm squire were arranged to be a square with an interval of 120 μm.


98

Fig. 1. Optical microscope image of fabricated 100-pixel 3D-STJ array.

2.2 Experimental setup The 100-pixel STJs were cooled to 0.31 K on a cold stage of a liquid-helium-free helium-3 cryostat. The bias point of the STJs was set near Δ/e, which was about 0.4 mV, by using a constant current source. A magnetic field of about 10 mT was applied along the diagonal direction of the junctions to suppress dc Josephson current. The output signals of the STJs were amplified using current amplifiers, then the pulse height spectra of the output signals were obtained by using FPGA-DSP based multi-channel analysers with trapezoidal pulse shaping of 10 μs. X-ray detection performance of the STJs was evaluated using fluorescent X-rays, which were generated using a X-ray tube equipped with a pure Al target and a carbon nanostructure electron emitter [9].


3.1 Current-voltage (IV) characteristics Fig. 2 showed a current-voltage (IV) curve of a typical 3D-STJ. A subgap current was about 40 nA and an energy gap was 0.44 meV. Ninety-three 3D-STJs in 100-pixel exhibited similar IV characteristics and were able to work as energy-dispersive X-ray detectors by the same bias current of 40 nA at the magnetic field of about 10 mT.


99

Fig. 2. IV curve of the typical STJ.

3.2 Soft X-ray detection characteristics An energy spectrum of the STJ for fluorescent X-rays from the pure Al target was displayed in Fig. 3. Double peaks for Al-Kα (1487 eV) showed X-ray absorption events in the top and bottom electrode of the STJ. The absorption efficiencies of each electrode for Al-Kα X-rays were 38 % and 9 %, respectively. The double peak problem can be solved by the top electrode with the thickness of more than 600 nm. The C-Kα (277 eV) and O-Kα (525 eV) X-rays shown in the figure were emitted from carbon and oxide on the surface of the Al target, respectively. The energy resolution values are 12 eV FWHM (C-Kα) and 13 eV FWHM (O-Kα), respectively. In addition, we can observe the Fe-Lα and Si-Kα X-rays generated from iron and silicon as impurities in the pure Al. The concentration of the impurities (iron and silicon) will be less than 0.5 %. The 3D-STJs of working pixels show similar energy resolving power for soft X-rays


100

Fig. 3. Energy spectrum of the STJ.

4. Conclusions We have developed the energy-dispersive X-ray detectors based on the 100-pixel array of the 3D-STJs and evaluated performance in the soft X-ray range. The typical 3D-STJ exhibited the IV curve having a subgap current of about 40 nA and an energy gap of 0.44 meV. Ninety-three 3D-STJs in the 100-pixel had similar IV characteristics. The 3D-STJ showed enough energy resolving power to detect soft X-rays. The energy resolution values for the O-Kα was 13 eV FWHM, which was about 4 times higher than that of the best SSD. An obvious degradation in the energy resolution wasn’t observed for 3D-STJs. It is concluded that the fabrication process of 3D-STJ was established for supplying the high performance X-ray detectors for light elements to applications for innovative structural materials.

Acknowledgement The authors thank to T. Adachi and the members of clean room for analog–digital superconductivity (CRAVITY) for the 3D-STJ fabrication. This work was supported by Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office and JSPS KAKENHI Grant Number 15K17495.

References [1] D.M. Schlosser, P. Lechner, G. Lutz, A. Niculae, H. Soltaua, L. Struder, R. Eckhardt, K. Hermenaua, G. Schaller,

F. Schopper, O. Jaritschin, A. Liebel, A. Simsek, C. Fiorini, and A. Longoni, “Expanding the detection efficiency of silicon drift detectors,” Nucl. Inst. Methods in Phys., 624(2), 270-276 (2010).

[2] S. Shiki, N. Zen, M. Ukibe, and M. Ohkubo, “Soft X-Ray Spectrometer Using 100-Pixel STJ Detectors for Synchrotron Radiation,” AIP Conf. Proc., 1185, 409-412 (2009).

[3] M. H. Carpenter, S. Friedrich, J. A. Hall, J. Harris. And R. Canter, “Development of Ta-based STJ X-ray detector arrays for synchrotron science,” J. Low Temp. Phys., 176, 222-227 (2014).

[4] M. Ohkubo, S. Shiki, M. Ukibe, N. Matsubayashi, Y. Kitajima, and S. Nagamachi, “X-ray absorption near edge spectroscopy with a superconducting detector for nitrogen dopants in SiC,” Sci. Rep., 2, 831-1-831-5 (2012).

[5] G. Fujii, M. Ukibe, S. Shiki, and M. Ohkubo, “Development of array detectors with three-dimensional structure toward 1000 pixels of superconducting tunnel junctions,” IEICE Trans. Electron, E98-C(3), 192-195 (2015).

[6] K. Hinode, S. Nagasawa, M. Sugita, T. Satoh, H. Akaike, Y. Kitagawa, and M. Hidaka, “Pattern-Size-Free Planarization for Multilayered Large-Scale SFQ Circuits,” IEICE Trans. Electron, vol.E86-C(12), 2511-2513 (2003).

[7] https://unit.aist.go.jp/riif/openi/cravity/en/index.html


101

[8] G. Fujii, M. Ukibe, and M. Ohkubo, “Improvement of soft X-ray detection performance in superconducting-tunnel- junction array detectors with close-packed arrangement by three-dimensional structure,” to be published in Supercond. Sci. Technol..

[9] H. Sugie, M. Tanemura, V. Filip, K. Iwata, K. Takahashi and F. Okuyama, “Carbon nanotubes as electron source in an x-ray tube,” Appl. Phys., Lett. 78, 2578-2580 (2001).


102

Atom Probe Analyses of Carbides on Prior Austenite Grain Boundaries in 9Cr

Heat Resistant Martensitic Steel

Byeong-Chan Suh1), Taisuke Sasaki1), Fujio Abe2), Kazuhiro Hono1) 1) Magnetic Materials Unit, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki, Japan

2) Materials Reliability Unit, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki, Japan +81-29-859-2718, +81-29-859-2701, [email protected]

Abstract: Trace additions of boron and nitrogen in 9Cr martensitic steel are known to improved creep strength; however, their role has not been understood very well because of the lack of characterizations of their alloying behaviour. We found boron and nitrogen are enriched in metal carbide and fine V/Nb-enriched particles using transmission electron microscopy and three dimensional atom probe. The V/Nb-enriched carbonitrides are formed in contact with the metal carbide along prior austenite grain boundaries. No segregations of boron and nitrogen are confirmed at the prior austenite grain boundaries. The partitioning of boron in metal carbide and V/Nb-enriched carbonitride is considered to make them thermally stable, resulting in a prevention for coarsening of metal carbide during further heat treatment, such as creep test or welding thermal cycle.

1. Introduction 9Cr martensitic steel strengthen by the addition of boron and nitrogen developed at National Institute for Materials Science (MARBN12) has improved creep strength of both base metal and welded joint [1]. Therefore, this steel comes into the spotlight as a promising candidate material for high temperature applications in the Advanced Ultra-Supercritical (A-USC) boiler. On the other hand, the strengthening mechanism by the addition of boron and nitrogen is controversial, due to the difficulty of detecting boron and nitrogen in the martensitic microstructure. Tabuchi et al. reported that segregation of boron on the grain boundaries decreases the grain boundary energy, which changes the mechanisms of the α-γ transformation during the weld thermal cycle, resulting in an improved creep strength [1]. On the other hand, Liu et al. reported that boron diffuses into the matrix after tempering 9Cr steel, resulting in an enrichment of boron in metal carbides, not segregation at the grain boundaries [2]. Therefore, the present study aimed at clarifying the partitioning and segregation behavior of boron and nitrogen in the tempered 9Cr martensitic steel by transmission electron microscopy and three dimensional atom probe.

2. Experimental procedure The chemical compositions of the 9Cr martensitic steel are shown in Table I. A steel plate was normalized at 1100 oC for 1 h followed by tempering at 800 oC for 1 h. Microstructure characterization was performed by scanning electron microscope (SEM), transmission electron microscope (TEM) and 3D atom probe (3DAP). TEM observation was carried out using Titan G2 80–200 (S)TEM with a probe aberration corrector. Thin foils for the TEM observation were prepared by focused ion beam (FIB, Helios Nanolab 650) followed by ion-milling using a Gatan precision ion polishing system (PIPS). 3DAP analyses or atom probe tomography (APT) were carried out with a locally built laser-assisted atom probe using a femtosecond laser pulse at a wavelength of 343 nm [3]. Square bars with dimensions of 1 x 1 x 20 mm3 were prepared by Helios Nanolab 650 and fine milled by Carl Zeiss Beam 1540EsB to prepare sharp and needle-like specimen. The 3DAP analyses were performed in an ultra-high vacuum condition (<1.0 × 108 Pa) at a base temperature of 20 K.

Table I. Chemical composition of present MARBN12 steel (wt.%)

Fe C Si Mn Cr W Mo Co Ni V Nb N O B Sol. Al Bal. 0.081 0.30 0.53 9.09 2.62 <0.01 3.02 <0.01 0.21 0.052 0.0072 0.003 0.011 0.001

3. Result and discussion Fig. 1a shows the secondary electron micrograph of tempered 9Cr martensitic steel. Prior austenite grain boundaries (PAGBs) are clearly distinguished as indicated by yellow broken lines. Fig. 1b and 1c show the high-angle annular dark-field (HAADF) STEM image and EDS elemental maps of Fe, Cr, Co, W, Mn, V and Nb, respectively. Cr, Mn and W are enriched in the metal carbide on the prior austenite grain boundary, which was identified as M23C6 particle by nanobeam electron diffraction. In addition, V and Nb are enriched along the PAGB. Thin V/Nb-enriched phase in less than 10 nm in thickness and 20 nm in length is observed in contact with the M23C6 along the PAGB.


103

Figure 1 (a) SEM and (b) HAADF-STEM images after tempering treatment in MARBN12 steel, and (c) EDS elemental

maps of the region in Fig. 1b.

A sharp needle-like specimen including a PAGB was prepared by FIB as shown in Fig. 2 (a). There is a particle in the vicinity of the tip apex, which is imaged with a brighter contrast in the back-scattered electron (BSE) image, as shown in Fig. 2a. Fig. 2b and 2c show the atom probe tomography of C, B, Nb and V obtained from this specimen. In the bottom part of the APT, a metal carbide enriched with C, B, Cr and W is observed. In addition, there are two V/Nb-enriched particles in contact with and slightly away from the metal carbide as shown in Fig. 2c.

Figure 2 (a) BSE image of sharp and needle-like specimen, and 3DAP maps of (b) C/B, and (c) V/Nb elements.

The chemical composition determined from the atom map, Fe30.4Cr44.9Mn0.9Co0.6W4.4C12.8B2.46, suggests

that the metal carbide is M23C6 in an agreement with the EDS result and the electron diffraction pattern. The average concentration in the α′ phase was determined to be Fe86.3Cr8.3Mn0.5Co2.8W0.4C0.14B0.01. Liu et al. also reported that the concentrations of carbon and boron in the M23C6 metal carbide were 18.3 and 2.0 at.%, respectively, in 9Cr steel with 0.11wt.% C, 0.001wt.% B and 0.05wt.% N [2]. Such a difference in the concentrations of metal carbide between this steel and present 9Cr martensitic steel may be attributed to a significant difference in the stability of the metal carbide.


104

There are two fine V/Nb-enriched particles formed in contact with and slightly away from metal carbide. The chemical composition of the V/Nb-enriched particle slightly away from metal carbide is Fe4.0Cr7.9Mn0.3Co0.09W4.6C10.0B0.08, and the V and Nb concentration of 14.8 and 29.3 at.%, respectively, are much higher than the nominal values (0.232 and 0.0315 at.%, respectively). In addition, there is also an enrichment of carbon and nitrogen. The concentration of carbon is about 10.0 at.%, which is slightly less than the carbon concentration of metal carbide (12.8 at.%) and much higher than its nominal value. Nitrogen is also enriched in V/Nb-enriched particles. The concentration of nitrogen is 25.2 at.%, which is much larger than its nominal value (0.0289 at.%). Therefore, such a V/Nb-enriched particle is considered as V/Nb carbonitride. It demonstrates that the addition of nitrogen strongly affects the formation of V/Nb-enriched particles. The concentration of all solute elements in V/Nb-enriched carbonitride formed slightly away from metal carbide is almost similar to V/Nb-enriched carbonitride formed in contact with the metal carbide. Similar V/Nb-enriched particles (MX nitrides) having a concentration of 29.11 at.% V, 15.59 at.% Nb and 32.64 at.% N were reported in 9Cr steel with 0.015 wt.% N and 0.01 wt.% B [4, 5], but it is different from V/Nb-enriched carbonitride observed in the present work.

Sharp needle-like specimen for three dimensional atom probe includes the PAGB, as indicated by yellow broken line in Fig. 2a. As mentioned above, V/Nb-enriched carbonitride are only observed along prior austenite grain boundaries. Therefore, the boundary including two distinct V/Nb-enriched carbonitrides, as shown as yellow line in Fig. 2c, would be the PAGB, as expected in BSE image (Fig. 2a). It has been reported that boron is segregated at prior austenite grain boundaries and that such a free soluble boron at prior austenite grain boundaries affects the strengthening in 9Cr steel [1]. However, no segregations of solute atoms were observed along the prior austenite grain boundary.

Metal carbides are formed at prior austenite grain boundaries and lath/packet boundaries. Boron is enriched in M23C6 metal carbides, thereby stabilizing the carbide. Fine V/Nb-enriched carbonitride are formed not only along prior austenite grain boundaries but also in contact with metal carbides. Especially, such V/Nb-enriched carbonitrides formed in contact with metal carbide would prevent metal carbide coarsening during further thermal treatments, such as creep test or weld thermal cycles.

4. Conclusion After tempering treatment, boron is enriched at M23C6 metal carbide along prior austenite grain boundaries with the concentration of about 2.5 at.%. In addition, fine V/Nb-enriched carbonitrides are formed in contact with the metal carbide and along prior austenite grain boundaries. Such an enrichment of boron in metal carbide and V/Nb-enriched nitrides is considered to suppress the coarsening of metal carbide.

References [1] M. Tabuchi, H. Hongo and F. Abe, "Creep Strength of Dissimilar Welds for Advanced USC Boiler Materials", Metall. Mater.

Trans. A, 45, 5068–5075 (2013).

[2] F. Liu, D.H.R. Fors, A. Golpayegani, H.O. Andrén and G. Wahnström, "Effect of boron on carbide coarsening at 873 K (600 °c) in 9 to 12 pct chromium steels", Metall. Mater. Trans. A, 43(11), 4053–4062 (2012).

[3] K. Hono, T. Ohkubo, Y.M. Chen, M. Kodzuka, K. Oh-ishi, H. Sepehri-Amin, F. Li, T. Kinno, S. Tomiya, Y. Kanitani, “Broadening the applications of the atom probe technique by ultraviolet femtosecond laser”, Ultramicroscopy, 111(6), 576-583 (2011).

[4] P. Hofer, M.K. Miller, S.S. Babu, S. a. David and H. Cerjak, "Investigation of Boron Distribution in Martensitic 9% Cr Creep Resistant Steel", ISIJ Int., 42, S62–S66 (2002).

[5] P. Hofer, M.K. Miller, S.S. Babu, S. a. David and H. Cerjak, "Atom probe field ion microscopy investigation of boron containing martensitic 9 Pct chromium steel", Metall. Mater. Trans. A, 31(13), 975–984 (2000).


105

3DAP/TEM study on Boron Partitioning Behavior in 9Cr-3Co-3W Heat-Resistant Steel

Taisuke Sasaki Magnetic Materials Unit, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki, Japan


106

6 MV Tandem Accelerator System for Ion Beam Analysis of Structural Materials at the University of Tsukuba

Kimikazu Sasa 1), 2), Akiyoshi Yamazaki 2), Shigeo Tomita 2), Masanori Kurosawa 3), Satoshi Ishii 1), Hiroshi Naramoto 1), Daiichiro Sekiba 1), 2), Tetsuaki Moriguchi 1), 2), Eiji Kita 1), 2)

1) Tandem Accelerator Complex, Research Facility Center for Science and Technology, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, 305-8577, Japan, phone: +81-29-853-2494, fax: +81-29-853-2565, email: [email protected]

2) Graduate School of Pure and Applied Sciences, University of Tsukua, 3) Graduate School of Life and Environmental Sciences, University of Tsukuba

Abstract: A new tandem electrostatic accelerator for ion beam analysis (IBA) was installed in the spring of 2014 at the University of Tsukuba. It consists of a 6 MV Pelletron tandem accelerator, five negative ion sources, and twelve beam courses for ion beam applictions. The 6 MV Pelletron tandem accelerator will be applied not only to IBA, but also to areas such as nanotechnology, accelerator mass spectrometry (AMS), heavy ion irradiation, and nuclear physics. Routine beam delivery and IBA experiments for structural materials will start in 2015.

1. Introduction The University of Tsukuba’s Tandem Accelerator Complex (UTTAC) is a major center of ion beam research in Japan. We have operated and maintained the 1 MV Tandetron accelerator, the 1 MV high-resolution Rutherford back scattering (RBS) system and the radio-isotope utilization equipment. We planned to install the new horizontal-type 6 MV Pelletron tandem accelerator in the experimental room on the first floor to replace the damaged 12UD Pelletron tandem accelerator by the Great East Japan Earthquake of 11 March 2011 [1]. A three-year plan for the new accelerator’s construction was started in 2012 [2]. The 6 MV Pelletron tandem accelerator will be used for various ion-beam research projects, such as AMS, microbeam applications, particle-induced X-ray emission (PIXE) analysis for geoscience and materials research, heavy ion RBS and elastic recoil detection analysis, nuclear reaction analysis for hydrogen in materials, and high-energy ion irradiation for semiconductor and nuclear physics.

2. Design of the 6 MV tandem accelerator system

All experimental equipment is installed on the first floor at UTTAC. Only rooms for control and data analysis are on the second floor. Fig. 1 gives an overview of the equipment on the first floor. The new accelerator system consists of the 6 MV Pelletron tandem accelerator, four new ion sources, the Lamb-shift polarized ion source (S1 in Figure 3), and five new beam courses. After the 105° analyzer magnet, the beam line is separated in two directions by the 40° switching magnet in the accelerator room.

2.1 Injector (ion sources) In Fig. 1, S2 is a high-current Cs-sputtering negative-ion source (SNICS II), and S3 is a radio frequency charge exchange ion source (Alphatros) to produce He− beams for injection into the 6 MV Pelletron tandem accelerator. S4 and S5 are the multi-cathode Cs-sputtering negative-ion sources (MC-SNICSs) for AMS.

At the low-energy side, the injection energy from the ion sources is 65 keV. There are three 90° retractable electrostatic spherical analyzers (ESAs) with a 200 mm radius and 35 mm plate separation to provide for each of the five ion sources. These analyzers will greatly reduce energy dispersion of the injected beam, and eliminate possible interference from the high- and low-energy tails of the abundant beams. The 90° high-mass-resolution double-focusing magnetic analyzer with a 457 mm radius has a mass energy product of ME/Z2 = 15 amu MeV with a mass resolution of up to 200. A 15 kV pulsed power supply is provided to bias the 90° magnetic analyzer chamber for the sequential injection of up to four different beams with a maximum mass difference of 17%. Downstream of the 90° magnetic analyzer, two movable offset Faraday cups are located to measure particle transmission through the accelerator and the beam intensities of lower-mass beams during higher-mass beam injection. In front of the accelerator tank, there is a beam attenuator to adjust the transport for high-beam current by decreasing one-tenth of the beam current.

2.2 Accelerator The main accelerator (model 18SDH-2 Pelletron accelerator developed by NEC, USA) is a dual acceleration electrostatic accelerator. The first model of this type was installed at the Australian Nuclear Science and Technology Organisation at almost the same time as the second model’s installation here at the University of Tsukuba. The accelerator tank is about 2.74 m in diameter and 10.5 m long. It was delivered to Tsukuba in March 2014. The high-voltage terminal has a long gas stripper tube assembly and a foil changer with 80 foil holders for equilibrium stripping ions. Carbon stripper foils will be used for heavy ions in a high charge state. The stripper gas canal is about 10 mm in diameter and 95 cm long. The terminal voltage can be as high as 6.3 MV. Stability is estimated to be better than 1 kV at


107

a 6.0 MV terminal voltage. Maximum beam currents are predicted to be up to 50 μA for heavy ions. Fig. 2 shows an energy range of the 6 MV tandem accelerator. It is possible to accelerate a wide variety of particles with a high-resolution energy value for ion beam applications.

Fig. 1. Overall view of the first floor at UTTAC. The 6 MV Pelletron tandem accelerator and its beam lines, the 1 MV Tandetron accelerator, and a positron-annihilation spectrometry system are set up on the first floor.

Fig. 2. Ion beam energy range for the 6 MV tandem accelerator.

Machine shop

10 m

1 MV Tandetron

Experimental Room

Accelerator Room

Experimental preparation room

Polarized Ion Source (PIS)

SF6 storage tank

6 MV tandem

A1

A2

A3

A4

A5A6

A7

L5

L4

L3

L2L1

Vertical line

Positron annihilation

MC-SNICSs

Mossbauer spectrometer

AMS

S1S2

S3 S4

S5

105º

40º

90º55º

Beam transport

0 10 20 30 40 50 60 70 80

H,DHeBCOF

AlSiClCaTi

CuBr

MoAg

ITaAu

H,DHe

BCOFAl

SiCl

CaTi

CuBr

MoAg

ITaAu

Total Beam Energy (MeV)

Ion


108

2.3 High-energy beam transport line The high-energy beam transport line after the accelerator has a mass energy product of ME/Z2 = 176 amu MeV. The 105° analyzer double-focusing magnet (1.27 m radius) has a resolution of M/ΔM = 725, and allows only ions with the proper charge state, mass, and energy to pass into the beam-measurement system. After the 105° analyzer magnet, three movable Faraday cups are set to intercept precisely the various beams of interest. A 90° deflection double focusing magnet (1.27 m radius and ME/Z2 = 176 amu MeV) is provided along the beam line. Two magnetic quadrupole triplet lenses are arranged to control divergence and focus ion beams between the 105° analyzer magnet and the switching magnet. The switching magnet is installed at the end of the beam line in the accelerator room, and has five beam ports.

Another high-energy beam transport line equipped with a vertical ion-irradiation system is connected from the accelerator room to the existing experimental room, which houses seven beam courses. A total of twelve beam courses will be available for ion beam analysis (IBA) of structural materials, nuclear physics, ion beam applications, and AMS.

2.4 Beam lines and IBA system In the accelerator room, the IBA system equipped with a high-precision four-axis goniometer is located on the L1 beam course. The L2 beam course has a large environmental testing chamber (1 m diameter) that will be mainly used for the radiation-resistant testing of semiconductor devices. The L3 beam course is constructed as follows: the Oxford microbeams quadrupole lens system [3] is used to obtain spot diameters of the order of 1 μm and below, and a superconducting detector [4] is used for high-sensitivity PIXE analysis. The L4 beam course is the rare-particle detection system for AMS. The L5 beam course is a general-purpose line for ion beam applications. In the experimental room, there are two large magnetic spectrographs (A6 and A7 beam courses) for nuclear physics. Other beam courses are used for ion beam channeling (A2 beam course), 3D fabrications with swift heavy ions (A3 beam course), and hydrogen analysis (A4 beam course). There is also a large general-purpose chamber (A5 beam course).

We have a plan to carry out ion beam analysis of structural materials at L1, L3 and L4 courses in the program of SIP-IMASM. Structural and light element analyses will be performed by the IBA system at L1 and L3 courses. Isotope trace experiments in structural materials will be planned and executed by using AMS techniques at L4 course. Fig. 3 shows the 6 MV tandem accelerator and the IBA system for structural materials.

(a) (b)

Fig. 3. Photographs of the 6 MV tandem accelerator. (a) A full view of the accelerator system. (b) An ion beam analysis equipment for material science at L1 course.

3. Conclusions The new 6 MV Pelletron tandem accelerator was installed at UTTAC in 2014, and regulated the beam transport. The new multi-purpose tandem accelerator will see various uses, including, IBA, AMS, ion irradiation, and nuclear physics. The micro-beam analysis system for light element in structural materials is currently under development at L3 course. In addition, development of the superconducting detector proceed in parallel for the micro-beam analysis system at National Institute of Advanced Industrial Science and Technology (AIST). It will start routine IBA measurements of structural materials in 2015.


109

Acknowledgements We would like to thank all our colleagues at the Tandem Accelerator Complex, University of Tsukuba, for their hard work in the post-quake recovery. We also acknowledge Hakuto Co., Ltd. Japan and National Electrostatics Corp., USA for installation support for the 6 MV Pelletron tandem accelerator and their dedicated effort.

This work is supported by the Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.

References [1] K. Sasa, “Damage Situation of the 12UD Pelletron Tandem Accelerator at the University of Tsukuba by the Great

East Japan Earthquake”, Proceedings of HIAT 2012, Joint Accelerator Conferences Website, 80–82 (2012).

[2] Kimikazu SASA, “The 6 MV tandem accelerator project for nuclear physics and ion beam applications at the University of Tsukuba”, AIP 1533, 184-188 (2013).

[3] G.W. Grime, F. Watt, “Focusing protons and light ions to micron and submicron dimensions”, Nucl. Instrum. Methods B 30 (1988) 227–234.

[4] M. Ohkubo, S. Shiki, M. Ukibe, N. Matsubayashi, Y. Kitajima, S. Nagamachi, “X-ray absorption near edge spectroscopy with a superconducting detector for nitrogen dopants in SiC”, Scientific Reports 2, article number:831 (2012).


110

Development of the Ion Microbeam System for Analysis of Light Elements in Structural Materials

at the University of Tsukuba

A. Yamazaki 1), K. Sasa 1,2), S. Ishii 2) , M. Kurosawa 3), S. Tomita 1), E. Kita 1,2) 1) Faculty of Pure and Applied Sciences, University of Tsukuba,

1-1-1 Tennodai, Tsukuba, Ibaraki 305-8577, Japan, +81-29-853-2498(phone) +81-29-853-2565 (fax),

[email protected] 2) Research Facility Center for Science and Technology, University of Tsukuba

3) Faculty of Life and Environmental Sciences, University of Tsukuba

Abstract: A new submicron scanning nuclear microprobe line is been constructing at the University of Tsukuba. The OM-2000 nuclear microscope end stage of the Oxford Microbeams Ltd. is installed to characterize structural materials. This ion microbeam system will be mainly used for X-ray imaging of two dimensional distributions for light elements in solid sample using particle induced X-ray emission (PIXE) technique. A silicon drift detector (SDD) with a thin window of Si3N4 is adopted for the measurement of light elements. In addition, a superconducting tunnel junction (STJ) array detector will be installed at the end stage for the purpose of more efficient measurement for the light elemental distribution in the near future.

1. Introduction Research and development of new structural materials are indispensable for industrial growth. It is well known that the characteristics of structural materials are strongly affected by containing light trace elements. Therefore, observation of the light trace elements in material is important for the development of the new structural materials. However, this technique is so difficult that many of the conventional methods of measurements are not applicable for observing light elements in material.

Particle induced X-ray emission (PIXE) is one of useful methods for elemental analysis with ion beam and is suitable for detecting trace elements because of its good signal-to-noise ratio. Combining a microbeam scanning technology, we can obtain a two-dimensional map of trace elements in a sample material. Therefore, this method (microbeam PIXE method) can be a powerful tool for R & D of new structural materials.

In the University of Tsukuba, a new 6 MV tandem accelerator have been installed, and now a construction of an ion microbeam irradiation line is proceeding in the tandem accelerator facility. Two X-ray detectors, a silicon drift detector (SDD) and a superconducting tunnel junction (STJ) array detector will be installed on the target chamber to detect soft X-rays from light nuclei in target materials efficiently. This paper describes the plan and the present status of the ion microbeam line of under construction in our facility.

2. Construction of an ion microbeam system Schematic illustration of the ion microbeam system in our facility is shown in Fig. 1. This beam line is going to connect to the 0 degree beam course after the switching magnet on the 6 MV tandem accelerator beam line. A submicron microbeam scanning end stage OM-2000 (Oxford Microbeams Ltd., UK) will be installed at the end of this

Fig. 2. Schematic illustration of the high resolution ion microbeam system at the University of Tsukuba.


111

microbeam line. This end stage consists of beam defining slits, a ferrite cored pre-lens deflector coil for X-Y scan, a triplet of magnetic quadrupole lenses and a target irradiation chamber with three axes target positioning device. The first slits (object slits) will be installed on a front part of the microbeam line. The distance from the first slits to the target position in the irradiation chamber is 8750 mm and the working distance (distance from the end of the last quadrupole magnet to the target position) is 165 mm. A photograph of the OM-2000 microbeam scanning end stage is shown in Fig. 2.

Fig. 3. Photograph of the microbeam scanning end stage OM-2000 (Oxford Microbeams Ltd., UK).

A silicon drift detector (SDD) is adopted for detecting characteristic X-ray emitted from the irradiated samples. It

has advantages of good energy resolution in high count rate condition and easy handling without liquid nitrogen courant. For the measurement of light elements such as B, C, N, a thin silicon nitride (Si3N4) film of 40 nm thick with an

aluminum coating of 30 nm thick is adopted as a front entrance window. Transmission for the characteristic X-ray of B is 19.7%. The diameter of the entrance window is 5 mm. Photographs of our SDD, XR-100FastSDD (Amptek Inc., US) with pulse processing unit and its front window is shown in Fig. 3(a) and (b), respectively.

Fig. 4. Photograph of the silicon drift detector XR-100FastSDD with digital pulse processor (Amptek Inc., US) (a). An entrance window is made of silicon nitride (Si3N4)with a thickness of 40 nm for observing characteristic X-rays from light elements (b).

In addition, our facility has a plan to install a superconducting tunnel junction (STJ) array detector [1] on the target chamber. This STJ array detector has been developed in National Institute of Advanced Industrial Science and Technology (AIST) to observe soft X-rays more efficiently with better energy resolution. In order to install the both X-ray detectors, a new large target chamber will be fabricated in the near future.


112

3. Summary and future plan We are constructing an ion microbeam irradiation system for analyzing light elements in structural material in our tandem accelerator facility at the University of Tsukuba. A silicon drift detector (SDD) and a superconducting tunnel junction (STJ) array detector will be installed on the target chamber aiming at efficient measurements of characteristic X-rays of light elements in the target material. The construction of the beam line and installation of OM-2000 end stage with the SDD will be finished in December 2015. Preliminary PIXE measurements with microbeam scanning conditions will start at the beginning of 2016. And the installations of the STJ array detector and the new target chamber will be completed until the end of March, 2015.

Acknowledgement The authors would like to thank technical staffs at the tandem accelerator facility for their valuable advice and various technical supports. This work is supported by Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM) operated by the cabinet office.

References [1] S. Shiki, M. Ukibe, Y. Kitajima, M. Ohkubo, “X-ray detector performance of 100-pixel superconducting tunnel

junction array detector in the soft X-ray region”, J. Low Temp. Phys., 167, 748-753 (2012).


113

In situ Observation of Reduction Kinetics of Iron Oxides

K. Kimijima1), Y. Niwa1), M. Kimura1) 1) Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK)

1-1 Oho, Tsukuba-shi, Ibaraki, 305-0801 JAPAN TEL +81-29-864-5200 (ex. 2547), E-mail: [email protected]

Abstract: Time-resolved measurement of the reduction process of α-Fe2O3 was performed using in situ-QXAFS technique. In order to study the mechanism of the reduction process of iron-oxides, measurements under different reduction gases were conducted. The reduction processes were successfully studied using in situ QXAFS spectra at the Fe K-edge during reduction reaction using different reduction gas. In both cases using H2 and CO, the reduction reaction of α-Fe2O3 involved two steps as previously reported: the first step was a very fast process in which FeIII was reduced to FeII, and the second step was slower than the first step reaction, in which FeII was reduced to Fe0 metal. In this study, the reaction rate constants for the second-step were obtained. The measurement was performed under the condition in which the reduction gas concentration was sufficiently high at the oxide surface. Hence, the rate constant obtained from this experiment is relevant to the rate of electron transfer from the reduction gas to the oxide.

1. Introduction In a steel making process, raw material, iron ore (iron oxide), is reduced to Fe metal by CO-based reduction gas and pig iron is produced in a reaction furnace called a “blast furnace”. Because 10,000 tons level of pig iron is manufactured in one blast furnace per day and the energy required for the reduction process is very large, information on the reaction mechanism is important to achieve higher efficiency of the process. On the other hand, the reaction that occurs in a blast furnace is a very complicated reaction consisting of multiple processes. To comprehend its overall behavior, it is essential to understand the elementary reactions that comprise the reaction. Therefore, various iron oxides were reduced with different reductant, and the reduction process was measured by in situ-QXAFS. 2. Experiment A sample was prepared by diluting α-Fe2O3 with BN in SUS sample holder. In situ flow cell reported previously [1] was used in consideration of continuity with previous experiments. Under He gas flow (100 sccm), the temperature was elevated from room temperature to 800 °C at 40°C min-1 and then to 900 °C at 10 °C min-1. After reaching 900 °C, reaction was initiated by introducing a reduction gas. Reduction gas, which is H2 or CO diluted to 20 vol% by He gas, was introduced into the reaction cell at 100 sccm. The Fe K absorption edge was measured by a transmission method using QXAFS method to obtain a spectrum at 20 s intervals. All XAFS experiments were carried out at the High Energy Accelerator Research Organization (KEK), Photon Factory (PF), BL-9C. 3. Results and discussion Figure 1a shows the spectrum changes of α-Fe2O3 under H2 gas atmosphere at 900°C. As the reduction reaction proceeds, the spectra of α-Fe2O3 (FeIII) → FeO (FeII) → Fe (Fe0) was observed. The reduction rate of FeIII to FeII is extremely fast, and the reaction is completed in about 2 spectrum measuring period of QXAFS (20 - 40 s). In the process of second-stage, FeII → Fe0, since the isosbestic point was observed in the spectrum, the reaction is considered to proceed without metastable intermediate phase. The reaction rate of FeII to Fe0 was estimated from the change in absorbance. To derive the apparent reaction rate constant, time change of absorbance (μt) at 7124 eV, which is the white line of α-Fe2O3 was obtained (Figure 1b). As the change in absorbance decreased exponentially, reduction of FeO is considered to be the first-order reaction with respect to Fe(II) concentration. Based on the fitting results of the plot in Figure 1b, the apparent reaction rate constant k was determined to be 3.8 × 10-1 ± 0.01 min-1.

On the other hand, when using CO as reduction species, as shown in Figure 2, it can be seen that the reaction proceeds very slowly compared to H2. Also in this case, because the attenuation curve of absorbance can be approximated by an exponential function (Fig. 2b), it can be regarded as going through a pseudo-first order reaction with respect to iron oxide concentration. Concentration of reduction species at the oxide surface, in other words, the difference in the diffusion rate to the oxide surface according to the type of gas, would not have a significant impact on the reaction rate, at least under the conditions of this experiment (reduction gas partial pressure, feed rate). The apparent reaction rate constant determined from the absorbance change was 4.0 × 10-2 ± 0.001 min-1.

It is considered that a reduction reaction of the iron oxide particles is a series of reactions including absorption of reduction species onto iron oxide surface, diffusion/reduction of oxygen atoms within the oxide (oxidation of reduction species), and desorption of oxidized/reduced species. It is unlikely that diffusion of oxygen atoms in the oxide is dependent on the difference in reduction species. In the case of this research, the reduction reaction at the iron oxide surface was found to be the rate-limiting processes. In this study, the measurement was performed under the condition,


114

in which the reduction gas concentration was sufficiently high at the oxide surface. Hence, the rate constant obtained in this report is relevant to the rate of electron transfer from the reduction gas to the oxide.

μ μ

Figure 1 (a) Time evolution of XANES spectrum at K edge of α-Fe2O3 under H2 (20 %) balanced with He (total flow rate 100 sccm), (b) the absorbance change according to the reaction time (at 7124 eV).

μ μ

Figure 2 (a) Time evolution of XANES spectrum at K edge of α-Fe2O3 under CO (20 %) balanced with He (total flow rate 100 sccm), (b) the absorbance change according to the reaction time (at 7124 eV). The experiment conditions were the same as those in Fig.1 except reduction gas.

4. Conclusions Reduction reaction mechanism of α-Fe2O3 was studied using different reduction gases (H2 and CO). In this study, apparent reaction rate constant for the reduction of oxide to metal was successfully obtained. For further studies, conditions of reduction gas on the oxide surface, such as concentration of reduction gas and temperature dependency, will be examined.

References [1] M. Kimura, Y. Uemura, T. Takayama, R. Murao, K. Asakura, M Nomura, “In situ QXAFS observation of the

reduction of Fe2O3 and CaFe2O4”, Journal of Physics: Conference Series, 430, 012074 (2013).


115

Characterization of Microelement in Ferritic Heat-Resistant Steels by TOF-SIMS

Norimichi Watanabe 1), Hiroaki Mamiya 1), Fujio Abe 2), Masataka Ohkubo 3), Hideaki Kitazawa 1) 1) Quantum Beam Unit, National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, JAPAN

TEL +81-29-851-3354, E-mail: [email protected] 2) Materials Reliability Unit, National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, JAPAN 3) Department of Electronics and Manufacturing, National Institute of Advanced Industrial Science and Technology (AIST)

1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, JAPAN

Abstract: It is well known that when the ferritic heat-resistant steel is doped with a small amount of boron, the grain boundary structure becomes stabilized, and the creep strength is improved. However, it has not been shown why the structure close to the grain boundary becomes stable in the boron-doped ferritic heat-resistant steel. In the present study, we observed the spatial distribution of boron on the surface of the ferritic heat-resistant steel using TOF-SIMS (time-of-flight secondary ion mass spectrometry). B ions were clearly detected in the positive mass spectrum. We made a direct observation of boron which is microelement in ferritic heat-resistant steel using TOF-SIMS. 1. Introduction

In TOF-SIMS, the surface of the sample is bombarded by the primary ions, and the secondary ions sputtered from the sample surface are separated by the time of flight [1]. We can obtain information about the elements and molecular species on the sample surface at a high detection sensitivity using TOF-SIMS. Furthermore, in-plane distribution of material components as well as a mass spectrum can be obtained in TOF-SIMS [2]. In this study, we observed the distribution of microelements in ferritic heat-resistant steels using TOF-SIMS. It is known that when boron is added to ferritic heat-resistant steel, the crystal grain boundary becomes stable and the creep strength is improved [3-4]. It is important to investigate the distribution of microelements to understand the mechanism in the characteristic improvement with added boron [5]. In reference [5], although BO2

- secondary ions were detected by use of Bi3+ primary

ions, B+ secondary ions were not directly observed. Therefore, we got interested in how we can observe the low-mass ions when we used Ga+ primary ions which can detect low mass at a high sensitivity. We have attempted to measure two-dimensional image of the distribution of B+ in ferritic heat-resistant steel using Ga+ primary ions, and investigate the correlation between the grain boundary and the boron distribution in the sample surface.

2. Experiment Figure 1 shows the principle of TOF-SIMS. In TOF-SIMS, pulsed primary ion bombards the sample surface and mass separation of the secondary ions sputtered from the sample surface is performed by measuring the difference of time of flight, as shown in Fig. 1. We used primary Ga+ ion beam of 30 kV and measured the secondary ions sputtered from the sample surface in an unbunching mode (high spatial resolution mode). In a bunching mode, the mass resolution is improved by compressing the pulse of the primary ion without reducing the pulse current because the mass resolution depends on the pulse width of the primary ion beam [6]. However, a beam spot of the primary ion is enlarged and spatial resolution is degraded in the bunching mode. Therefore, we used the unbunching mode to observe the local distribution of microelement in the vicinity of the grain boundary with the higher spatial resolution in this measurement. Fig.1 Principle of TOF-SIMS

Primary ion: Ga+

Sample detector

Secondary ion

Fig.2 Experimental apparatus (TOF-SIMS)


116

Figure 2 shows experimental apparatus of PHI TRIFT V nanoTOF (Ulvac-Phi Inc.). The sample submitted for the measurement is 9Cr-3W-3Co-0.2V-0.05Nb-0.08C steel with 150 ppm boron [5]. Polishing and buffing of the sample surface was performed. Furthermore, the surface was sputtered by the argon gas gun before the SIMS analysis to clean the surface. The grain boundary was observed after the sputter cleaning.

3. Results and Discussion Figure 3 shows the typical positive secondary ion spectrum of the heat-resistant steel after the Ar sputtering. Several secondary ions generated by the bombardment of the metal surface were detected in the positive mass spectra. The peak of B+ secondary ions is also clearly observed.

Fig. 3 Positive mass spectrum in ferritic heat-resistant steel (a) Fe+, (b) Cr+, (c) Mn+, (d) B+

55.31 55.95 56.38 50.90 52.09

54.59 54.93 55.23 10.98 11.01 11.04

Mass Mass

Mass Mass

(a) (b)

(c) (d)

Fe+ Cr+

Mn+ B+

Fig. 4 Two-dimensional images of total ions, Cr+, Fe+, Al+, Mn+, B+ in ferritic heat-resistant steel using TOF-SIMS. (Scan area: 100μm × 100μm)

20 μm

20 μm


117

In addition to the mass spectrum, we measured two-dimensional images of total ions, Fe+, Cr+, Al+, Mn+, B+ in ferritic heat-resistant steel using TOF-SIMS, as shown in Fig. 4. The B+ was locally-distributed in the vicinity of the grain boundary. Furthermore, Mn+ and Al+ were also observed in the same area as B+. We were able to clearly observe the spatial distribution of B+ on the surface of the ferritic heat-resistant steel using TOF-SIMS.

4. Conclusions We were able to measure B+ secondary ions in the positive mass spectrum and in-plane B+ distribution in the surface of ferritic heat-resistant steel using TOF-SIMS. We made a direct observation of B+ which is locally distributed in ferritic heat-resistant steel. We obtained the new evidence that the boron exists in the Mn and Al. We plan to investigate the correlation between the boron distribution and the crystal grain boundary in samples with different boron concentration and samples after creep tests.

Acknowledgement This work is supported by SIP (Cross-ministerial Strategic Innovation Promotion Program)-IMASM (Innovative measurement and analysis for structural materials).



[2] Bruno W. Schueler, “Microscope imaging by time-of-flight secondary ion mass spectrometry”, Microsc. Microanal. Microstruct., 3, 119-139 (1992).

[3] T. Horiuchi, M. Igarashi, F. Abe, “Improved Utilization of Added B in 9Cr Heat-Resistant Steels Containing W" ISIJ Int., 42, S67 (2002).

[4] F. Abe, T. Horiuchi, K. Sawada, "High-Temperature Annealing for Maximization of Dissolved Boron in Creep-Resistant Martensitic 9Cr Steel", Mater. Sci. Forum., 426-432, 1393 (2003).

[5] S. Suzuki, R. Shishido, T. Tanaka, F. Abe, ”Characterization of the Inhomogeneous Distribution of Light Elements in Ferritic Heat-Resistant Steels by Secondary Ion Mass Spectrometry”, ISIJ Int., 54 (4), 885 (2014).

[6] P.K. Dutt, “Theory of “ion-bunching” in relation to the development of an electrostatic time-of-flight mass spectrometer”, Nucl. Instrum. Methods., 10, 37 (1961).


118

Poster papers

IMASM Theme 3 Heterogeneous Boundaries


119

Interface Characterization of Al/Co Laminated Film on a Si substrate using Variable Temperature TOF-SIMS

Norimichi Watanabe 1), Hiroaki Mamiya 1), Masataka Ohkubo 2), Hideaki Kitazawa 1) 1) Quantum Beam Unit, National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, JAPAN

TEL +81-29-851-3354, E-mail [email protected] 2) Department of Electronics and Manufacturing, National Institute of Advanced Industrial Science and Technology (AIST)

1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, JAPAN

Abstract: We have analyzed the sample surface of various materials using TOF-SIMS (time-of-flight secondary ion mass spectrometry).In this study, we installed a heating stage in TOF-SIMS to investigate the diffusion phenomenon of light element and microelement in the interface between an environmental barrier coating film and a metal. We have performed TOF-SIMS measurement for the Al (25nm)/Co (2nm) laminated film as a test sample in the temperature range of 25 to 580 that is above eutectic point 575 of Si-Al alloy. We successfully observed the change of the structure in the interface between the Si substrate and Al film in a hot environment. (Paper :Submit by August 31, 2015.) 1. Introduction In TOF-SIMS, pulsed primary ion bombards the sample surface, and the mass spectrometry of atoms and /or molecules can be carried out with high mass and high special resolution by measuring the time until the sputtered secondary ions reach a detector [1]. A phenomenon that light secondary ion reaches a detector earlier than heavy secondary ion is utilized for the mass spectrometry. The information about elements or molecular species within depth of about 1 nm below the sample surface is obtained with high detection sensitivity. We have evaluated various materials such as ferritic heat resistant steels for a high-efficient thermal power plant and structural materials for an aircraft using TOF-SIMS. We introduced the heating stage which can change the sample temperature from the room temperature to 600 into a TOF-SIMS to evaluate the change of elements in the materials under high temperatures. For instance, the study of diffusion process between the environmental barrier coating (EBC) and the heat resistant alloy in turbine blades of an aircraft engine is one of our reasons for introduction of the heating stage in TOF-SIMS. In this study, we fabricated the Al (25nm)/Co(2nm) laminated film on the Si(100) substrate by magnetron sputtering system as a test sample, and we observed the structure of Al/Si interface in the room temperature and 580 using TOF-SIMS.

2. Experiment We used the Ga liquid metal ion gun for the primary ion of TOF-SIMS. The secondary ions generated by the irradiation of the primary ions are analyzed by the time-of-flight mass analyser [2], and it becomes possible to investigate what kind of atom and molecule is on the sample surface. Although TOF-SIMS obtains the information of atomic species within depth of about 1 nm below the sample surface, we can measure three-dimensional image of the sample surface by analysing and sputtering the sample surface alternately using argon gas gun other than primary ion gun [3-5]. We improved the conventional sample stage, and developed the heating stage, as shown in Fig. 1. We set the R type thermocouples and the nichrome wire heater into a stainless steel block.

Fig. 1 Heating stage of TOF-SIMS. The figure on the right is Al/Co films on the Si (100) substrate which is put on the SUS 304 stage. The sample is covered with a mesh to fix the sample.

Thermocouple

Heater Hot plate stage Al/Co film


120

The temperature is controlled by a proportional-integral-derivative (PID) controller. We measured Al (25nm)/Co (2 nm) films deposited by magnetron sputtering system on the Si (100) substrate as test measurement under the high temperature environment. We measured depth profiling of Al/Co film on the Si substrate and three-dimensional image of interfacial reaction in Si/Al interface.

3. Results and Discussion Figure2 shows time-temperature characteristics in the heating stage of TOF-SIMS. The red line shows temperature and the black line shows heater power in Fig. 2. In this experiment, the temperature was increased to 580 . It takes about 20 minutes for temperature to be stable when it increases from 500 to 600 , as shown in Fig. 2. We measured the Al (25nm)/Co (2nm) laminated films on the Si (100) substrate using TOF-SIMS. The area of the surface analysis is 100 μm ×100 μm and the area of the sputtering for three-dimensional measurement is 400 μm ×400 μm. We analyze the surface in the area where the surface is sputtered. Because the area which is sputtered is much wider than the area which is analyzed, the area of the surface analysis is almost flat. Figure 3 shows the intensity of Al and Co as a function of sputtering time in the temperature of 25 and 580 . In the temperature of 25 , the top coated Co film was removed after the surface was sputtered for 5 seconds because the intensity of Al increased and saturated and the intensity of Co decreased after the sputtering process for 5 seconds, as shown in Fig. 3. There was not significant change between Al and Si in the temperature of 25 . On the other hand, the change was observed at the 30 seconds in the temperature of 580 .

Figure 4(a) shows three-dimensional image measured in the temperature of 25 . Each layer of the Si substrate, Al and Co layer was clearly observed. Figure 4(b) shows three-dimensional image measured in the temperature of 580 . We were able to confirm that Si penetrated Al layer in Fig. 4(b). Since the eutectic point of Al-Si alloy is 575 [6], a solid-liquid coexistent phase should exist when the temperature is above 575 . Therefore, the complicated structure between Co and Si must be caused by the fact that the sample temperature is exceeded over the eutectic point.

Fig. 2 Time-temperature characteristics in the heating stage of TOF-SIMS

Fig.3 TOF-SIMS depth profiling of Al/Co film on Si (100) substrate

200

400

600

0 0 2000 4000 6000

10

20

30

0

Time [sec] Sputtering time [sec] 0 20 40 60

Fig. 4 Three-dimensional image of Al/Co film on Si (100) substrate using TOF-SIMS measured at (a) 25 , (b) 580 . The analyzed size in the x-y plane is 100 μm×100 μm.

(a) (b)

Co

Al

Si

Co

Si

Coexistence of Si and Al


121

4. Conclusions We installed the heating stage which covers the temperature ranging from the room temperature to 600 into TOF-SIMS. We measured three-dimensional image of Si/Al interface using TOF-SIMS as a test of the heating stage, and observed the change of the structure in the Si/Al interface around 580 . We plan to observe the diffusion phenomenon of light element and microelement in the interface between an oxide coating film and a metal using variable temperature TOF-SIMS.




[2] Bruno W. Schueler, “Microscope imaging by time-of-flight secondary ion mass spectrometry”, Microsc. Microanal. Microstruct., 3, 119-139 (1992).

[3] J.S. Fletcher, N.P. Lockyer, S. Vaidyanathan, J.C. Vickerman, “TOF-SIMS 3D Biomolecular Imaging of Xenopus laevis Oocytes Using Buckminsterfullerene (C60) Primary Ions”, Anal. Chem., 79(6), 2199–2206 (2007).

[4] G.L. Fisher, A.M. Belu, C.M. Mahoney, K. Wormuth, N. Sanada, “Three-Dimensional Time-of-Flight Secondary Ion Mass Spectrometry Imaging of a Pharmaceutical in a Coronary Stent Coating as a Function of Elution Time”, Anal. Chem., 81(24), 9930–9940 (2009).

[5] Michael A. Robinson, Daniel J. Graham, David G. Castner, “ToF-SIMS Depth Profiling of Cells: z-Correction, 3D Imaging, and Sputter Rate of Individual NIH/3T3 Fibroblasts”, Anal. Chem., 84(11), 4880–4885 (2012).

[6] M.M. Makhlouf, H.V. Guthy, “The aluminum–silicon eutectic reaction: mechanisms and crystallography”, J. Light Met., 1(4), 199 (2002).


122

Precipitation induced variations in mechanical and magnetic properties for X-750 superalloy

H. Mamiya1), J. Rabajczyk1,2), N. Watanabe1), A. Kowalska1,2), H. Kitazawa1,2), 1) Quantum Beam Unit, National Institute for Materials Science, Sengen 1-2-1, Tsukuba 305-0047, Japan

2) Faculty of Materials Science and Engineering, Warsaw University of Technology, Woloska 141, Warsaw 02-507, Poland [email protected]

Abstract: For a nickel-chromium-based superalloy, Inconel X-750, we confirmed that precipitation occurs during aging heat treatments from 933 to 1013 K. Furthermore, we found for the first time that the precipitation not only mechanically hardens the superalloy but also induces soft magnetic response.

1. Introduction

Nickel-chromium-based superalloy, Inconel X-750, shows high mechanical strength and resistance to creep at high temperatures, hence, it has been used for rotor blades and wheels, bolts, and other structural members in turbine engines. It is also used extensively in rocket-engine thrust chambers. Another feature of this superalloy is its high resistance to chloride-ion stress-corrosion cracking. It is typically used for internal components in nuclear plants. Furthermore, this superalloy has excellent strength and ductility at cryogenic temperatures. Therefore, it has been used cryogenic storage tanks and tubes of superconducting wires. Considering such wide range of applications from ultra-low to ultra-high temperatures, a comprehensive study on the nature of this superalloy from various perspectives at various temperatures is highly desired.

In this superalloy, it has been well known that the solid solution decomposes to Ni3Al based γ′ precipitates with Ll2-ordered structure and γ matrix with disordered fcc structure below about 1200 K. With the precipitation and their growth during heat treatments below the temperature, called aging, the high-temperature strength is significantly improved, because the precipitates impede the movement of dislocations. For this reason, effects of the nano-structural evolution on mechanical property have been intensively investigated over the last several decades. On the other hand, little attention has been given to its magnetic property, although this superalloy has been used as tubes of superconducting wires in high magnetic fields at ultra-low temperatures. To our knowledge, diffuse cluster superparamagnetism was discussed only in one article. Important issues such as the relationship between the nano-structural evolution and variation of the magnetism have been untouched, although the information on the magnetic properties is helpful not only in advancing cryogenic applications but also in providing a new nondestructive inspection method for the quality of the rotor blades. In this paper, we report effects of aging on structural, mechanical and magnetic properties of X-750 nickel-chromium-based superalloy in order to complement the partially lacking knowledge.

2. Experimental

Nickel-chromium-based superalloy, X-750, was supplied from Special Metals in the sheet form of 1.0 mm of thickness. The chemical composition is as follows: Ni 70.68 %, Cr 16.21 %, Fe 8.49 %, Ti 2.55 %, Nb 0.93 %, Al 0.65 %, Mn 0.21 %, Co 0.10 %, Si 0.09 %, C 0.05 %, Cu 0.02 %, Ta 0.01 %, P 0.004 % and S 0.001 %. The samples were solid solution heat treated at 1423 K for 1 hr, and then water-quenched, followed by an aging at the temperatures Ta from 933 to 1013 K for 2 hrs. After air cooling, the samples were polished to remove the damage introduced in these heat treatments.

Structural characterization was performed by using a conventional optical microscope and small-angle X-ray scattering (SAXS) camera (NanoSTAR, Bruker AXS.) For optical microscopy, the samples were etched in Marble's reagent after mechanical polishing. The samples for the SAXS measurements were mechanically polished, typically to 30 μm, to achieve the appropriate transmission rate. To exclude the effects of thickness variation, the thickness of the samples was determined using the X-ray transmission rate and line absorption coefficient calculated from the chemical composition of the samples. Using typical Cu Kα X-ray radiation, the SAXS measurement of superalloys is limited due to low transmission rates. Hence, we used Mo Kα radiation (the wavelength λ is 71 pm) focused using a Goebel mirror.

Mechanical properties were measured using a dynamic ultra-micro-hardness tester (DUH-211, Shimadzu Co.) with a Berkovich indenter. The indentation was performed inside the grains with a working load of 100 N. Magnetic properties were observed using a SQUID magnetometer (MPMS, Quantum Design.) The samples were the same thin foil as SAXS to reduce the demagnetizing factor Nd to 0.03 or less. Magnetization curves were measured in the magnetic field range |H| < 4 MA/m at various temperatures between 2.0 and 300 K. The temperature dependences of


123

magnetization M were measured using the following protocols: Briefly, the zero-field-cooled magnetization MZFC was measured on a heating process in H = 80 A/m with constant temperature sweep rate of 1.0 K/min after the samples were cooled to 2.0 K without applying magnetic field; the field-cooled magnetization MFC was recorded in H = 80 A/m on a cooling process from a paramagnetic state at 300.0 K, where the sweep rate is 1.0 K/min.


Optical microscopy shows that the grain size of γ matrix increases from 70 μm to 150 μm as Ta is raised from 933 to 1013 K. SAXS intensity of the as-solution-heat-treated sample simply decreases in inverse proportion to the fourth power of q, where the modulus of scattering vector q is given by 4πsinθ /λ and θ is half the scattering angle. On the other hand, the q-dependent profile of the intensity for the samples with aging treatments shows a clear hump whose location shifts from q = 0.7 nm−

1 to 0.3 nm−1 as Ta is raised from 933 to 1013 K. The analyses using McSAS fitting code

indicate that there are precipitates with the average radius of 3, 4, 5, 6, and 7 nm in the samples with aging treatments at Ta of 933, 953, 973, 993, and 1013 K, respectively, whereas no precipitates exist in the as-solution-heat-treated sample. Thus, we can confirm that the growth of the precipitates is accelerated with increasing Ta.

Ultra-micro-hardness test shows that Berkovich micro-hardness increases from 450 to 630 in the samples with aging treatments as Ta is raised from 933 to 1013 K, while the hardness of the as-solution-heat-treated sample is approximately 260. In other words, the precipitation hardens X-750 as known well. On the other hand, the magnetic measurements show the magnetic susceptibility of the as-solution-heat-treated sample shows a typical spin-glass-like behaviour at the temperatures below 18 K. In contrast, the magnetic susceptibility for the samples with aging treatments steeply increases with decreasing temperature from almost 60−70 K and exhibits a diverging tendency around Tw ∼ 40 K. Then it keeps much high magnitude below Tw. We can say that this soft magnetic property is correlated with the precipitates, because Ni3Al becomes ferromagnetic below 60−70 K [1]. In brief, a precipitation induced soft magnetic property is found in a nickel-chromium-based superalloy, Inconel X-750, for the first time.

4. Concluding remarks

For a nickel-chromium-based superalloy, Inconel X-750, we confirmed that the precipitation occurs during the aging treatments from 933 to 1013 K. Furthermore, we found for the first time that the precipitation not only mechanically hardens the superalloy but also induces soft magnetic response. This knowledge is useful in the design for X-750 tubes of superconducting wires.


References [1] H. Mamiya, M. Demura, and H. Kitazawa, "Aging and rejuvenation in a ferromagnetic Ni3Al", Eur. Phys. J. B 88,

114 (2015).


124

Advanced In Situ Multi-functional Characterization of High Strength CFRP Materials

Hongxin Wang1), Hideaki Kitazawa 1), Masamichi Kawai 2), Daisuke Fujita 1)

1) National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Tel +81-29-859-2000, Fax +81-29-859-2801, [email protected],

2) University of Tsukuba,

Abstract: The carbon fiber reinforced plastics (CFRP) is a plastic matrix composite material reinforced by carbon fibers. In our research, we aimed to realize the best performance of CFRP for the aircraft by investigating the relationship between the structure and mechanical properties of CFRP. High resolution image of CFRP was obtained by helium ion microscopy firstly. The structure of CFRP showed plastic matrix covered around carbon fibers. Then 2D micro characterization of CFRP with externally applied stress was analyzed by confocal Raman microscopy. The composition of CFRP could be obtained. In order to clarify the role of interfaces between carbon fibers and plastic matrix for high strength of CFRP, mechanical properties of high strength in CFRP was analyzed by scanning probe microscopy.

1. Introduction Since the carbon fiber reinforced plastics (CFRP) has a high strength-to-weight ratio and a high rigidity, it is widely used for technical applications including aerospace and automotive. However such composite materials are composed of complex multiphase systems of heterogeneous materials. The mechanical behaviors of composites are closely related not only to in situ properties of components including the matrix and the carbon fibers, but also to characteristics of the interfaces between the matrix and the carbon fibers. Due to the microscopic dimensions of the components and the interfaces, it is difficult for the traditional testing methods to measure the mechanical properties of the corresponding regions in the microstructures. Nanoindentation technique using a standard microscopic indenters is an effective approach to measure in situ microscopic mechanical properties of heterogeneous materials. The system of conventional nano-indenters combined with nano-mechanical atomic force microscopy (AFM) has realized the true nanoscopic measurement of mechanical testing. It enables to obtain in situ mechanical properties in microscale domains and the 3-D indentation morphology simultaneously [1-3].

2. Experimental Before the materials characterization, CFRP samples were cut to rectangular sheets. After the three-point or four-point bending tests, the sample surfaces were polished by using a cross section polisher (CP, IB-09020CP). The surface morphology was observed by Helium ion microscope (HIM, Orion plus, Carl Zeiss) with nanoscale resolution. Surface damage property was characterized by confocal Raman microscopy (Renishaw). Microscale hardness was characterized by so-called nano-indentation testing (TI950 Triboindenter, Hysitron). Nanoscale mechanical properties were characterized by atomic force microscope (AFM, Multimode 8, Bruker AXS) set in the glove box.

3. Results and Discussion Figure 1 shows the cross section images of the CFRP material using HIM. The structure of CFRP showed plastic matrix covering the carbon fibers. Typical diameter of carbon fibers was about 7 μm. The surface of carbon fiber was found to be rough although the surface of CFRP was polished by CP.

Fig. 1. HIM image of CFRP.

Fig.2. Bending test of CFRP sheets (a,b,c). HIM image of fractured CFRP.


125

Figure 2 shows the bending test of CFRP rectangular sheets. The typical dimension of CFRP sample is 27 mm in length, 1 mm in width and 1 mm in thickness. Elastic and plastic deformation occurred on high strength CFRP. The morphological distribution of carbon fibers existed in the corresponding region of the cross section. It meant that when the surface of CFRP was flat, the cross section of CFRP was flat. The rough carbon fiber surface can be seen rough cross section.

Figure 3 shows Raman spectrum of a carbon fiber of CFRP. The Raman spectrum of a carbon fiber is characterized by the presence of two peaks, corresponding to D-band and G-band, respectively. Comparing the Raman spectra among them, there was no obvious peak shift after the fracture caused by the bending test. Due to the content or polarizability of carbon fibers, the intensity changed. By analyzing the peak intensity of D-bands induced by defects and G-bands, the observed D/G ratio was 90 %, indicating the intrinsic defective property of carbon fibers. Figure 4(a) shows the AFM image of CFRP after conventional micron-scale indentation test. Although the marks of a triangular pyramid were formed on a carbon fiber, they were not clear enough to obtain the real area even with loading force of 4.5 mN, which was the maximum of the indenting apparatus. Figure 4(b) and (c) showed the AFM images of a carbon fiber and matrix surface after AFM nano-indentation. The nanoscale holes formed on the carbon fiber (b) and on the matrix (c) by a diamond AFM tip were found to be enlarged with the increase of applied forces up to ∼μN range. Figure 5 shows the dependence of the measured area of nanoholes formed on the CFRP by changing the loading force. Hardness can be approximately evaluated as the loading force divided by the nano-hole area. Inverse of the slope in Fig. 5 is the hardness. In this case, the observed hardness of the carbon fiber and the matrix is 925 MPa and 42 MPa, respectively.

4. Conclusions We have been developing the in situ characterization techniques at the nanoscale and micron scale for the high strength composite materials. The effectiveness of the characterization methods including Raman and Helium ion microscopies with bending test and the in situ nanoscale hardness characterization using AFM is demonstrated on the CFRP materials.

Acknowledgement This work was supported by Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative

Measurement and Analysis for Structural Materials (SIP-IMASM).

References [1] Q. Zhao, Y. Liang, K. Cheng, S. Dong, “Investigation of AFM-based nano-indentation on micro-machined silicon

surface”, Micronanoelectronic Technology, 1671-4776, (2003).

[2] L. Zhou, Y. Yao, “Single Crystal Aluminum Hardness Experiment Based on Nanoindentation and Atomic Force Microscopy (AFM)”, Mechanical Science and Technology, 25(1), (2006).

[3] X. Gao, Q. Yang, Z. Liu, “In situ characterization of carbon fiber/epoxy composites by nanoindentation”, Acta Material Composite Sinica, 29(5), (2012).

Fig.4. AFM images of CFRP after the conventional micron scale indentation (a), and the AFM nanoindentation (b)&(c).

Fig.5. Dependence of the area of nanoscale indentation by AFM on the CFRP with the applied loading force.

1000 1500 2000

500

1000

1500

Fig.3. Raman spectra of initial (a) and fractured

carbon fibers (b,c) fibers of CFRP.


126

3000 K Class Heating Action of Pulsed Electric Currents for High-Resolution

Transmission Electron Microscopy of High Melting Point Metals

Tokushi Kizuka, Yasuchika Suzuki, and Satoshi Murata

Division of Materials Science, Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba 305-8573, Japan, [email protected]

Abstract: We performed 3000 K class heating by application of pulsed electric currents in a transmission electron microscope. The method was capable of melting of even high melting point metals, e.g., tantalum. This achieving temperature enables studies of 1700 K class high-temperature resistant materials for harsh conditions.

1. Introduction Advanced materials have been used in various ultimate environments. The optimum textures of the materials in such environments are revealable by direct observations in the reproducible conditions, i.e., in situ transmission electron microscopy (TEM) [1,2]. The materials for structural applications, e.g., engine and aircraft materials, have been used at high temperatures exceeding 1700 K. Although high-temperature specimen stages for in situ TEM have been developed to study such heat-resistant materials, the stage temperature has still been limited under 1200 K for usual thin film type TEM specimens. In this study, we performed 3000 K class heating of a high-melting temperature metal, i.e., tantalum (Ta), by application of pulsed electric currents.

2. Methods The method used was based on in situ high-resolution TEM for experimental mechanics of materials combined with electric conductance measurements [3-9]. A schematic of this method is shown in Fig. 1. First, two Ta plates were prepared for use as electrodes. One of the plates was attached to the front of a cylindrical piezoelement on the first specimen holder for TEM. Then, another plate was attached to the second specimen holder for TEM. The contact edge of the plate was thinned to 5–20 nm. Both the holders were then inserted into the in situ TEM for nanometer manipulation at the University of Tsukuba. The specimen chamber of the microscope was evacuated, first by a turbomolecular pump and then by an ion pump, resulting in a vacuum of 1 × 10−5 Pa. Inside the microscope, piezomanipulation was used to contact the two tips. Pulsed currents were applied between the two plates to heat the contact. During this process, structural dynamics was observed in situ by lattice imaging via high-resolution TEM using a video capture system. The time resolution of image observation was 40 ms. Electrical conductance was measured by a

Fig. 1 In situ TEM for experimental mechanics of materials combined with electric conductance measurements.


127

two-terminal method. The results from high-resolution imaging and signal detection of conductance were simultaneously recorded and analysed for each image. By this method, structural dynamics was directly observed. Simultaneously, electrical behaviour, e.g., current, conductance, current–voltage characteristics, and electromigration, were also analysed. The size and shape obtained from the structural observations allowed estimates for conductivity and current density. In addition, the tensile tests of contacts could be performed using piezomanipulation. Therefore, the method enabled the investigation of high-temperature tensile tests of the materials under atomistic structural observations. 3. Results and discussion

3.1 Heating exceeding the melting point Figure 2 shows a time sequence of high-resolution images of a Ta contact formed between the two tips before and after application of a pulsed current. The lattice images of (110) of Ta was observed before the application of the pulsed current, as indicated by the arrow in Fig. 2(a), showing that the initial structure of contact had a crystalline structure. After the application of the pulsed current, the contrast of the contact changed to random, implying that an amorphous structure was formed. Thus, it is inferred that first, the crystalline contact was heated to melt by the current and subsequently was quenched. This is because quench was caused by rapid thermal transfer from the contact region to two semi-infinite reservoirs, i.e., the two tips [10]. Therefore, we could estimate the elevated temperature of the contact to be higher than the melting point of Ta, i.e., 3300 K. The result demonstrates that 3000 K class high-temperature states of heat-resistant materials are realized and can be investigated at the atomic resolution by this method. 3.2 High-temperature tensile tests Figure 3 shows a time sequence of high-resolution images of a tensile elongation process of a Ta contact while application of pulsed currents. In this case, the contact was tensiled by the manipulation of the upper tip along the direction indicated by the bold arrow in Fig. 3(a). As a result, the contact elongated by this tensile action, as shown in Fig. 3(b). The direction of the lattice fringes on the elongated region was parallel to that of the lower tip; the crystal orientation of these regions was the same. Also note that an amorphous structure emerged around the boundary between the upper tip and the elongated region. The results shows that the tensile force accelerated the crystallization after melting along the orientation of the neighboring tip. This observation corresponds to a 3000 K class high-temperature tensile test of a heat-resistant material.

4. Conclusion We performed heating of a high melting point metal, i.e., Ta, by application of pulsed electric currents. The method was capable of melting of Ta contacts. Thus, the heating temperature exceeded 3300 K, This achieving temperature enables study of 1700 K class high-temperature resistant materials for harsh conditions, such as engine and aircraft materials. By separation of the two tips, high-temperature tensile tests of heat-resistant materials can be performed.

Fig. 2 Time sequence of high-resolution images of a Ta contact formed between the two tips before and after application of a pulsed current.

Fig. 3 Time sequence of high-resolution images of a tensile elongation process of a Ta contact while application of pulsed currents


128

Acknowledgement This study was performed as a part of Cross-ministerial Strategic Innovation Promotion Program – Innovative measurement and analysis for structural materials. References [1] P. Hirsch, A. Howie, R. Nicholson, D. W. Pashley & M. J. Whelan. [Electron microsocpy of thin crystals], 2nd

Revised edn, (Krieger Publishing Company, 1977). [2] G. Dehm, J. M. Howe & J. Z. (eds.). [In-situ Electron Microscopy: Applications in Physics, Chemistry and

Materials Science], (Wiley-VCH Verlag &Co.KGaA, 2012). [3] T. Kizuka & N. Tanaka. "Dynamic high-resolution electron microscopy of diffusion bonding between zinc oxide

nanocrystallites at ambient temperature". Phil. Mag. Lett. 69, 135–139, (1994). [4] T. Kizuka, K. Yamada, S. Deguchi, M. Naruse & N. Tanaka. "Cross-sectional time-resolved high-resolution

transmission electron microscopy of atomic-scale contact and noncontact-type scannings on gold surfaces". Phys. Rev. B 55, R7398–R7401, (1997).

[5] T. Kizuka. "Atomistic visualization of deformation in gold". Phys. Rev. B 57, 11158–11163, (1998). [6] T. Kizuka. "Atomic process of point contacts in gold studied by time-resolved high-resolution transmission electron

microscopy". Phys. Rev. Lett. 81, 4448–4451, (1998). [7] T. Kizuka, H. Ohmi, T. Sumi, K. Kumazawa, S. Deguchi, M. Naruse, S. Fujisawa, S. Sasaki, A. Yabe & Y.

Enomoto. "Simultaneous observation of millisecond dynamics in atomistic structure, force and conductance on the basis of transmission electron microscopy ". Jpn. J. Appl. Phys. 40, L170–173, (2001).

[8] T. Kizuka. "Atomic configuration and mechanical and electrical properties of stable gold wires of single-atom width". Phys. Rev. B 77, 155401–155411, (2008).

[9] T. Kizuka & S. Ashida. "Free-space nanometer wiring via nanotip manipulation". Scientific Reports 5, 13529, (2015).

[10] L. Zhong, J. Wang, H. Sheng, Z. Zhang & S. X. Mao. "Formation of monatomic metallic glasses through ultrafast liquid quenching". Nature 512, 177–180, (2014).


129

Development of 1300 K Class Heating Stages for Transmission Electron

Microscopy of Thin Films

Tomo-o Terasawa and Tokushi Kizuka Division of Materials Science, Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan,

[email protected]

Abstract: A new type of a heating stage for transmission electron microscopy was developed for the studies of advanced materials used at high temperatures exceeding 1200 K, which is the limit of commercial heating stages. It was found that the heating temperature of our stage could be controlled to be higher than 1300 K. It is plausible that the further modification of the heater components increases the stage temperature, leading to the application of our method to the studies of engine and aircraft materials. 1. Introduction Transmission electron microscopy (TEM) has contributed to the progress of materials science because the method provides all the kinds of the information of microstructures, i.e., crystal structures, textures, compositions, surfaces, interfaces, grain boundaries, and point defects. In particular, in situ TEM reproduces the dynamics of microstructures in various environments in which materials are used. Since high-temperature environments are subjects to advanced structural materials, heating stages for in situ TEM have been developed [1, 2]. Various structural dynamics relating to texture control, e.g., recrystallization, phase transition, precipitation, and dislocation movement, have been investigated, resulting in the design of materials textures [3, 4]. However, the stage temperature has still been limited under 1200 K for usual thin film type TEM specimens [2, 3]. Thus, engine and aircraft materials used at the temperatures higher than 1700 K have been left out of the scheme. In this study, we have developed a new type of a heating stage for the studies of advanced materials used in such ultimate environments. 2. Experimental To observe the textures of bulk materials by TEM, the materials are polished to prepare thin films with a thickness less than 1 μm. The typical size of the observation area is a circular form of 3 mm in diameter. In addition, the pole piece gap of the objective lens, in which the specimen and heater are inserted, is at most several millimeters. We designed the heater satisfying these conditions on the basis of a Joule-heating method. The schematic illustration of the present experimental system is shown in Fig. 1. The heater was mounted on a current controlled specimen holder of a transmission electron microscope (JEOL JEM-2010KZ). As a “thermometer” of the heater, we deposited iron (Fe) thin films on the heater by argon ion beam sputtering at room temperature in Gatan DuoMill Model 600. The deposition was carried out for 2 min at the total pressure below 1 × 10−4 Pa. Although the temperature of the heater cannot be observed directly in the TEM, the phase transition is expected to suggest the specimen temperature.

Fig. 1. Schematic illustration of the system to observe heating behavior.


130

Before the specimen holder was inserted into the microscope, the heating states were checked in a vacuum chamber

in which a turbo-molecular pump achieved the total pressure below 1 × 10−4 Pa. The temperature was evaluated by a pyrometer through a viewport of the vacuum chamber. After the specimen holder was inserted into the microscope, the heating of the stage was carried out. The vacuum around the specimen and heater resulted in 1 × 10−5 Pa at the start of observation. After the heating, the structural dynamics of the specimens was observed in situ via high-resolution TEM using a video capture system. The accelerating voltage of electrons was 200 kV. The time resolution of image observation was 40 ms. 3. Results Figure 2 shows the photographs of the heater at various temperatures in a preliminary test in the vacuum chamber other than the microscope. The thermal radiation in a visible light region was enhanced as the bias voltage increased. We evaluated the heater temperature by a pyrometer from room temperature to 1310 K. Figure 3 shows the almost proportional increase in the heater temperature with respect to the bias voltage. It was thus confirmed that the heater temperature could reach at least 1300 K.

Fig. 2. Photographs of the heater in a vacuum chamber. The bias voltage in each photograph (a–c) is 0, 1.3, and 2.2

V while the heater temperature is room temperature, 973, and 1310 K, respectively.

Fig. 3. Relationship between bias voltage and heater temperature.

Inside the microscope, we observed the variation in the states of the Fe thin films to evaluate the heater temperature.

Besides the TEM observation, we took the photographs of the heater by a CCD camera, which indicated the strong thermal radiation from the heater as well as the preliminary test at 1300 K in the vacuum chamber. Figures 4(a) and 4(b) show the TEM images of the heater with the Fe thin films before and after the heating, respectively. The direct magnification of the images was 50,000. Before the heating, the edge of the heater was roughly coated by the Fe thin films. After the heating, the thin film aggregated on the heater surfaces, resulting in the formation of particles. This drastic change due to the melting of the thin films suggested that the temperature of the specimen was successfully


131

increased higher than the melting point of Fe, i.e., 1808 K although it exceeded the temperature range of the preliminary test. In addition, we can attempt further modifications of the components to increase the stage temperature toward higher levels.

Fig. 4. TEM images of the heater coated by Fe thin films.

4. Conclusions The new type of the heating stage for in situ TEM was demonstrated. The heating temperature was confirmed to be 1300 K which is higher than that of commercial heating units. Inside the microscope, the structural change of the Fe thin films deposited on the heater was observed after the heating. This change suggested that the heater temperature was successfully increased by our Joule-heating method. It is plausible that the temperature increases by further modification of the heater components, leading to the application of our method to the studies of engine and aircraft materials. Acknowledgement This study was performed as a part of Cross-ministerial Strategic Innovation Promotion Program – Innovative measurement and analysis for structural materials. References [1] T. Kamino, T. Yaguchi, M. Tomita and H. Saka, “In-situ high-resolution electron microscopy study on a surface reconstruction of

Au-deposited Si at very high temperatures”, Philos. Mag. A75, 105–114 (1997). [2] T. Kamino, T. Yaguchi, T. Sato and T. Hashimoto, “Development of a technique for high resolution electron microscopic

observation of nano-materials at elevated temperatures”, J. Electron Microsc. 54, 505–508 (2005). [3] N. P. Young, M. A. van Huis, H. W. Zandbergen, H. Xu, and A. I. Kirkland, “Transformations of gold nanoparticles investigated

using variable temperature high-resolution transmission electron microscopy”, Ultramicroscopy, 110, 506–516 (2010). [4] T. Yahiro and Y. Takai, “In-situ crystal structure analysis of cobalt nanocompounds synthesizing graphite at high temperatures”,

Jpn. J. Appl. Phys. 50, 015103 (2011).


132

Poster papers

IMASM Theme 4 Vacancy Defects


133

B. E. O’Rourke1*, L. Jiang1 R. Suzuki1 and N. Oshima1 1 National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial

Science and Technology (AIST), Tsukuba, Ibaraki 305-8568, Japan. *corresponding author, [email protected]

Abstract: At AIST the high intensity, slow positron facility is an electron accelerator based facility for measurement of positron lifetimes with intense, energy variable, slow-positron beams. There are two main beamlines, each of which has a capability to perform positron annihilation lifetime spectroscopy (PALS) with a standard large diameter (~ 10 mm) beam, or with a focused microbeam. The microbeam device is called ‘positron probe micro-analyzer’ (PPMA) and has a lateral resolution of around 50 μm. For both PALS and PPMA the energy of the positron beam can be varies from around 1 – 30 keV, corresponding to typical implantation depths of several nm to several μm. Using the penetration depth and lateral resolution capabilities of the AIST PALS and PPMA, we plan to characterize the defect distributions in structurally relevant materials as part of the SIP-IMASM project.

1. Introduction

Positron annihilation lifetime spectroscopy (PALS) is a powerful techniques to study the defect characteristics of materials [1-3]. Using a low-energy (0 – 30 keV), mono-energetic positron beam it is possible to control the penetration depth from the surface to a depth of several microns. Slow positron beam based PALS is thus used for the characterization of thin films, membranes, layered devices etc. Production of intense, slow positrons is complicated by the moderation process, in which energetic positrons are implanted into, then subsequently re-emitted into vacuum with a very low energy (~ eV) from a material with a negative work function for positron emission. This process has a very low efficiency, typically around 10-4 in the case of tungsten which is used at AIST. While radioisotope based positron sources are commercially available their intensity is limited by the activity of the source so that slow positron beams based on such sources have a limited intensity. More intense positrons beams can be generated using electron accelerators and nuclear reactors [4-7]. At AIST, a linear electron accelerator (LINAC) has been used to produce a positron beam for more than 20 years [8, 9]. In recent years a positron microbeam apparatus, the positron probe micro-analyzer (PPMA), has been developed [10]. Using the PPMA we have extended the functionality of the PALS method to 2D and 3D mapping of defect distributions [11] and to “in air” evaluation via the extraction of focused positron beams through extremely thin vacuum windows [12]. In the SIP-IMASM project we plan to use the functionality of the AIST slow positron facility to characterize the near surface of structurally relevant materials.

2. The AIST Facility

The accelerator based slow positron beam facilities at AIST are shown in Fig. 1. The electron beam from a 40 MeV electron LINAC can be directed to either of two production targets. In both targets the electron beam is extracted into air through a thin Ti window and is incident on a water cooled 5 mm thick Ta disk, the converter. Positrons are produced in the converter by irradiation of electrons from the LINAC, which typically provides an electron beam with a pulse width of 1-3 μs, a pulse rate of up to 85 Hz and an estimated maximum average electron current on the converter of a few μA. The moderator assembly is located directly behind the converter, inside a separate vacuum chamber, and is composed of a series of thin strips of tungsten films (thickness ~ 50 μm) arranged in a lattice like structure. The distance between adjacent strips is 5 mm and the total size of the assembly is a circle of diameter 30 mm. The moderator was assembled and then annealed at more than 2000 °C via irradiation with an intense electron beam in a separate vacuum chamber [13]. Slow positrons are extracted from the moderator by applying a small accelerating voltage (~ 10 V) and guided along the beam line duct by a uniform axial magnetic field supplied by the surrounding Helmholtz coils. The short positron pulse is then trapped in a linear storage trap by quickly increasing the potential of the input electrode after the pulse has passed. The output electrode is then slowly decreased so that a ~ ms long pulse is produced. This pulse is then chopped and bunched in order to produce a train of very short pulses (pulse length ~ 150 ps) at high frequency (equal to the chopper frequency, ~ MHz) incident on the sample [14].


134

LINAC (40 MeV)Electron Beamline

New Positron Beamline

Positron Beamline

SCA Electron Beamline

SuperconductingAccelerator (SCA)

PALS & PPMA

PAESPALS #2

PPMA #2

Convertor/Moderator 1

Convertor/Moderator 2

Electron Beamline

Positron Beamline

10 m

Electron Accelerator

Fig. 1. Layout of the AIST slow-positron facility. An electron LINAC can supply a 40 MeV electron beam to either of two positron production targets. One of the targets supplies the original positron beamline which has both the PALS and PPMA devices on the same beamport. A newly installed second beamline has separate ports for PALS and PPMA arranged in a vertical configuration. In future, it is planned to use a superconducting accelerator (SCA) in combination with the newly installed beamline for the 2nd generation, intense slow-positron beam facility. 2.1 Positron Annihilation Lifetime Spectroscopy (PALS)

At the PALS station, a beam with a diameter of around 10 mm is incident on the sample surface with an energy which

is variable between ~ 1 – 30 keV. Typically a series of samples (up to 5 or so) with sizes around 15 x 15 mm square are arranged on a single holder and placed in the sample chamber. This chamber is then pumped down to a vacuum of ~ 10-

6 Pa before opening the main beamline valve and starting the measurement. In the original PALS system the samples are interchanged manually by sliding the sample holder into position whereas in the new beamline the interchange can be automated and controlled by computer.

Depending on the analysis conditions and time constraints several million counts per lifetime spectra are obtained. Typical counting rates are around several 1000 cps so that a several spectra may be obtained in per hour. These spectra are then analysed by suitable software programs and the lifetime components extracted. In metals the lifetime is very sensitive to vacancy defects and the lifetime increases with increasing defect size (i.e. vacancy clusters as compared to mono-vacancies).

2.2 Positron Probe Microanalyzer (PPMA)

The PPMA focusses the positron beam to create a microbeam with a diameter of around 50 μm. In the original

beamline the PPMA is connected directly to end of the PALS system and takes the pulsed PALS beam and focusses it onto a transmission type remoderator [10]. The remoderator is made from a thin (~200 nm) film of Ni and has a moderation efficiency of around 10 %. It is used to increase the brightness of the beam as the re-emitted positrons have a low angle of divergence. The re-emitted beam is then accelerated and focussed again onto the sample. The impanation energy and time resolution are very similar to the PALS system with just a reduction in intensity by a factor of 10 or more due to the losses at the remoderator. As the PPMA is connected directly to the PALS system on the original beamline it takes a few days to switch between the two modes of operation. In the newly installed beamline facility, there are separate beamports for PALS and PPMA so that it is easy to switch between them [15].

The sample for PPMA is placed directly inside the sample chamber. This chamber is mounted on an x-y stage so that it is possible either to place many small samples on a single holder or to scan along a larger sample to map the defect distribution. A typical example of a map of defect distribution obtained by scanning the beam over a large sample is shown in figure 2 [16]. At each point on the map the positron beam should stay focussed on that point for as long as it takes to obtain sufficient counts for analysis. As mentioned in the previous section, for a full lifetime analysis typically several million counts may be collected. If such a number of counts were collected at each point on a large map then the time required would be prohibitive. Instead, typically a much simpler analysis, involving the determination of the


135

average lifetime (rather than a multi-component analysis) is performed. Such an analysis is possible with only a few thousand counts per spectra [16]. Under the SIP-ISAM project we plan to develop new software to efficiently perform this analysis and automatically produce a map of lifetime distributions in the quickest possible time.

Fig. 2. PPMA image, consisting of about 8 x 103 pixels, of a ductile fractured iron sample. The lifetime spectra at each

pixel contains around 5000 counts allowing the average lifetime to be determined. The total measurement time is around 1 day. Taken from ref. 16.

3. Future plans

In order to decrease the required measurement time we are exploring new ways to improve the intensity of the slow positron beam. This problem is particularly acute for mapping measurements at the PPMA where long measurement times are required. In the SIP-ISAM project the KEK group is developing methods to more efficiently pulse stretch and transport the slow positron beam [17]. The KEK developments are reported in these proceedings. At AIST we developing a next generation accelerator facility based on a superconducting accelerator (SCA) [18, 19]. A SCA has many advantages over a normal conducting LINAC, in particular regarding the pulse structure and duty cycle. For an SCA, very short pulses (~ ps) can be accelerated with very high rep. rates (MHz) with very high duty (up to cw in principle). This pulse structure is very advantages for slow positron beams as such beams may be transported and bunched without the need for pulse chopping. SCA can also be used to produce very short pulses of gamma radiation (Bremsstrahlung) which can be irradiated into a large sample in order to produce electron-positron pairs in-situ. The positrons annihilate in the bulk sample and by detecting the annihilation gamma rays information on positron lifetime can be obtained (the short gamma ray pulse providing the start signal). In combination with CT methods it should be possible to make a 3-D map of lifetime distributions in large (mm) scale samples.

4. Summary

The accelerator based slow positron facility as AIST provides intense, energy-variable positron beams for characterization of the near surface of materials via the positron annihilation lifetime spectroscopy (PALS) technique. Both standard, wide beam (~ 10 mm), and PALS with a microbeam (~ 50 μm), i.e. PPMA is available. Using these facilities in the SIP-IMASM project we plan to characterize the near surface of structurally relevant materials like iron and steel. In particular we are interested in defects induced both during sample preparation, i.e. cutting and polishing etc., and those defects which develop due to applied stress, hydrogen embrittlement, creep etc. The characterization of these defects in relation to the structural properties of the material can help with the identification of weaknesses before the material fails.

Acknowledgement

This work was supported by the Cabinet Office strategic innovation creative program innovative structural materials (SIP).

References [1] R. Krause-Rehberg, H. S. Leipner, Positron Annihilation in Semi-conductors (Springer-Verlag, Berlin, 1999)

[2] P. Coleman (Ed.), Positron Beams and their applications,


136

[3] P. J. Schultz and K. G. Lynn, Rev. Mod. Phys. 60 701 (1988).

[4] R. Krause-Rehberg, M. Jungmann, A. Krille, B. Werlich, A. Pohl, W. Anwand, G. Brauer, M. Butterling, H. Büttig, K. M. Kosev, J. Teichert, A. Wagner and T.E. Cowan, J. Phys. Conf. Series 262 012003 (2011).

[5] C. Hugenschmidt, G. Kögel, R. Repper, K. Schreckenbach, P. Sperr, B. Strasser and W. Triftshäuser, Nucl. Instrum. Methods Phys. Res., Sect. B 221 160 (2004).

[6] H. Schut, A. van Veen, J. de Roode, and F. Labohm, Mater. Sci. Forum 445–446 507 (2004).

[7] B.E. O’Rourke, N. Hayashizaki, A. Kinomura, R. Kuroda, E.J. Minehara, T. Ohdaira, N. Oshima and R. Suzuki, Rev. Sci. Instrum. 82 063302 (2011).

[8] T. Akahane, T. Chiba, N. Shiotani, S. Tanigawa, T. Mikado, R. Suzuki, M. Chiwaki, T. Yamazaki and T. Tomimasu, Appl. Phys. 51 146 (1990).

[9] R. Suzuki, T. Mikado, M. Chiwaki, H. Ohgaki and T. Yamazaki, Appl. Surf. Sci. 85 87 (1995).

[10] N. Oshima, R. Suzuki, T. Ohdaira, A. Kinomura, T. Narumi, A. Uedono and M. Fujinami, J. of Appl. Phys. 103 094916 (2008).

[11] N. Oshima, R. Suzuki, T. Ohdaira, A. Kinomura, T. Narumi, A. Uedono and M. Fujinami, Appl. Phys. Lett. 94 194104 (2009).

[12] W. Zhou, Z. Chen, N. Oshima, Kenji Ito, B. E. O’Rourke, R. Kuroda, R. Suzuki, H. Yanagishita, T. Tsutsui, A. Uedono and N. Hayashizaki, App. Phys. Lett. 101 014102 (2012).

[13] A. Yabuuchi, N. Oshima, H. Kato, B. E. O’Rourke, A. Kinomura, T. Ohdaira, Y. Kobayashi and R. Suzuki, Jap. J. App. Phys. Conf. Proc. 2 011102 (2014).

[14] R. Suzuki, Y. Kobayashi, T. Mikado, H. Ohgaki, M. Chiwaki, T. Yamazaki and T. Tomimasu, Jpn. J. Appl. Phys. 30 L532 (1991).

[15] B. E. O’Rourke, N. Oshima, A. Kinomura, R. Suzuki, Jpn. J. App. Phys. Conf. Proc. 2 011304 (2014).

[16] N. Oshima, R. Suzuki, T. Ohdaira, A. Kinomura, S. Kubota, H. Watanabe, K. Tenjinbayashi, A. Uedono and M. Fujinami, J. Phys. Conf. Sers. 262 012044 (2011).

[17] K. Wada et.al, These Proceedings

[18] B. E. O’Rourke, N. Oshima, R. Kuroda, R. Suzuki, T. Ohdaira, A. Kinomura, N. Hayashizaki, E. Minehara, H. Yamauchi,Y. Fukamizu, M. Shikibu, T. Kawamoto and Y. Minehara, J. Phys. Conf. Sers. 262 012043 (2011).

[19] B. E. O’Rourke, N. Oshima, A. Kinomura, T. Ohdaira and R. Suzuki, Mat. Sci. Forum 733 285 (2013).


137

L. Jiang*, B.E. O’Rourke, N. Oshima, and R. Suzuki 1 National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial

Science and Technology (AIST), Tsukuba, Ibaraki 305-8568, Japan. Phone number: +81 29 861 2469, Email address: [email protected]

Abstract: We investigated the near surface defect layer of stainless steel (SUS 316L) induced by electrical discharge machining (EDM) by using several analyzing methods including electron microscopes, x-ray diffraction and positron annihilation lifetime spectroscopy which were suitable for characterizing nano-structures in materials. By applying these analytical methods to the samples which were electrochemically polished for removing surface layer with well controlled thickness after EDM, we evaluated the depth dependence of induced defect.

1. Introduction

The need for high-performance structural materials with improved structural characteristics, e.g. high strength, light weight, increased heat resistance etc., is felt in many industries such as automotive, aviation, aerospace, where such materials can dramatically improve performance[1,2]. Generally, mechanical performance (and/or lifetimes) of structural materials are affected by their atomic scale / nano-scale structures. Therefore, in order to help develop high-performance structural materials and to continue the safe operation of existing structures, it is important to evaluate the materials nano-structure. In SIP-IMASM, we will use positron annihilation spectroscopy (PAS) together with other analytical methods to evaluate the nano-structure of structural materials for further understanding the relationship between nano-structure and mechanical-performance.

PAS is known to be a unique and powerful method for evaluating atomic-scale open volume defects in various materials [3]. Therefore PAS has been used to study deterioration of structural materials induced by fatigue [4-6], hydrogen-embrittlement, creep and cutting or preparation [7-10]. The principles of PAS is based on the fact that (i) positrons search and effectively trap in open volume defects such as vacancies and micro-voids because of the absence of an electrically repulsive positively charged nucleus, and (ii) the positron annihilation parameters (positron annihilation lifetimes spectrum and/or Doppler broadening spectrum of annihilation radiation) are strongly dependent on the environmental structure at the positron annihilation site, including defect size and density [11,12].

In order to accurately interpret the results of PAS (and/or other analytical methods), it is necessary to carefully consider the near surface defects additionally induced by the cutting process during sample preparation. There are reports about the study of positron annihilation on saw cutting [8-10], laser cutting, abrasive water cutting, and traditional milling cutting [7] indicating that the induced defect layer thickness is much larger than the penetration depth of positrons. We plan to use electro discharge machining (EDM) for preparing our steel sample in SIP. EDM is a well-established machining technique for manufacturing geometrically complex or hard material parts that are extremely difficult to machine by conventional machining processes and has long been employed in the automotive, aerospace, mold, tool and die making industries [13]. In the present study, we evaluated the defect layer of stainless steel (SUS316) induced by EDM using positron annihilation lifetime spectroscopy (PALS) and other characterization techniques such as X-ray diffraction (XRD) and electron microscope analysis . By applying these analytical methods to the samples which were electrochemically polished for removing surface layer with well controlled thickness after EDM, we evaluated the depth dependence of induced defects. SUS 316L is a suitable material to characterize the deformed layer induced by EDM for the first step experiments. The SUS316L, the low carbon version of type 316, remains a single phase austenitic material under all application conditions [14]. Therefore it is not necessary to consider effect of phase transformation in our data analysis.

In this paper, we report one part of our obtained results.

2. Experimental details 2.1 Sample preparation

Blocks of SUS316L with dimensions 14 mm x 14 mm x ~30 mm and specified chemical compositions of 0.011 mass%. C, 0.44 mass% Si, 0.79 mass% Mn, 0.020 mass% P, 0.001 mass% S, 17.11 mass% Cr, 12.24 mass% Ni, and 2.03 mass% Mo was obtained from the manufacturer. Firstly, the surfaces of the SUS316L blocks were polished with sand paper and washed in ethanol to remove surface impurities. The blocks were then annealed for 1 hour at a


138

temperature of 1100oC under vacuum, and then the vacuum tube containing the block was quenched in iced water [14,15]. Air was leaked into the vacuum tube after 10 minutes cooling in the ice water. This procedure dissolves carbide clusters [14] and other defects induced during sample preparation [7]. The blocks of SUS316L were then sliced by EDM to produce numerous samples with size 14 mm × 14 mm and thickness of 2 mm. EDM was performed in deionized water at temperature of 24 oC using a 0.2 mm diameter brass cutting wire. The maximum voltage and maximum current during electric discharge were 20~30 V and ~200 A, respectively. The duration time of one discharge was ~1 μs, therefore the heat energy in one discharge was a few mJ. The cutting speed and discharge frequency were ~ 4.0 mm/min and ~35 kHz, respectively. After producing the 2 mm thick sample slices by EDM, a portion of the sample surface layers was removed by electrochemical polishing. The solution used for this electrochemical polishing consisted of 100 mL of ethylene glycol 99.5% dissolved in 100 mL of distilled water, 10 g of NaCl, and 4.66 g of citric acid [16]. Only one side (area = 14× 14 mm2.) of each sample was polished by covering the other side with isolation tape. The total current was fixed to 0.4 A during polishing, and hence, the average current density was ~0.002 A/mm2 under the assumption that sample surface is flat. By varying the polishing time, samples with polishing depth in the range 0 ~ 100 μm were obtained. The polished thickness was measured by a micrometer (Teclock, SMD-565J-L) and step meter (KLA-Tencor, Alpha-step IQ). The micrometer has a probe size of ~ 10 mm and a typical resolution of less than 3 µm. The typical resolution of the step meter was less than ~0.1 nm for the z direction and ~ 1 μm for x-y plane direction. It should be noticed that the actual resolution for all directions are much broadened because the finite size of the probe (Diamond stylus ~5 µm). 2.2 Electron microscope analysis 2.2.1 Electron probe microanalyzer (EPMA)

The chemical composition and a cross-sectional image of near surface of SUS316L were evaluated by EPMA (JOEL, JXA-8500F) where the accelerating voltage is 15 kV, the beam diameter is ~10 nm, and the special resolution for EPMA is ~ 1 μm. The sample for cross section observation was prepared without electrochemical polishing by ion milling methods (JEOL, SM09010). 2.2.2 Energy dispersive x-ray spectroscopy (EDX)

The chemical composition of samples with various polished thickness was also evaluated by EDX (Hitachi-hightech, EMAX300) with an accelerating voltage of 20 kV. The EDX measurement was repeated for samples with various polished thickness.

2.3 XRD analysis

X-ray diffraction spectra were measured for samples with polished depth of 0 ~ 100 μm (Rigaku, AutoMATE). The X-ray implantation depth (Cr Kβ radiation) was around 10 μm. The size of beam-defining collimation slit of this system was 2.0 mm × 2.0 mm. Residual stress was measured in two perpendicular directions on the surface plane. 2.4 PALS analysis

For the positron lifetime measurement, a 22Na based positron lifetime spectrometer with BaF2 scintillators and a digital oscilloscope was used (see Fig. 1). The time resolution (FWHM) of the lifetime spectrometer was ~170 ps. The positron source, i.e., 22Na enveloped in the 7-μm kapton foil was sandwiched between two similarly prepared samples. About 3~6 ×10 6 annihilation events were accumulated for each lifetime spectrum at a counting rate of ~50 cps. The positron penetration depth was estimated to be ~30 μm.


139

Fig. 1. Schematic of the PALS system.

3. Results and Discussion

3.1 Results of electron microcope analysis

3.1.1 Results of EPMA analysis

Figure 2 (a) shows a cross-sectional image (magnification 1500) of the near surface of an un-polished sample. The cross section was prepared by ion milling methods. Figure 2 (b) shows the data obtained by the step meter for the un-polished sample. From these images we know that typical roughness with typical maximum amplitudes from peak to valley is ~ 10 μm. By using data in Fig. 2(b) the roughness given as standard deviation was about 1.5 µm. Figure 2 (c) shows a cross-sectional image (magnification 3000) which is one part of enlarged image of Fig. 3(a). The red line is the line scanned in the EPMA measurement and the result is shown in Fig. 2(d). The position denoted as z = 6.2 μm in Fig. 2(c) and 2(d) is the top surface of the sample. The EPMA signals changed markedly from z = 6.2 μm to 9.7 μm (see Fig. 2(c)). However, some part of this change may be due to existence of voids (see Fig. 3(d)), although it might be true that chemical content was changed by EDM. In addition, EPMA signals changed gradually further several micro-meters from z = 9.7 μm where scanned surface seems to be flat (see Fig. 3(d)). Those facts indicate that chemical distribution was changed by EDM with typical depth of about 10 μm. It should be noticed that Cu was detected from near the surface to z = 15 μm (see Fig 2(c)). Because the original SUS316L does not include Cu, we conclude that the Cu is deposited from the brass wire during EDM.


140

Fig. 2. (a) A cross-sectional image (magnification 1500) of the surface of an EDM cut SUS316L sample. (b) The roughness of the SUS316L sample as evaluated by a step meter. (c) A cross-sectional image (magnification 3000) of the surface of an EDM cut SUS316L sample (d) EPMA elemental composition as measured by scanning the electron beam along to the red line is shown in Fig. 2(c).

3.1.2 Results of the EDX analysis

We also evaluated the elemental compositions of an EDM cut SUS316L sample which was then electrochemically polished by EDX in order to evaluate a much deeper area ( ~ 50 um). Fig 3 (a) shows a photograph of one such sample in which different small regions were polished for 1, 2, and 5 min. The polished thickness as measured by a step meter were ~11, 20, and 42 μm respectively. Figure 3 (b) shows the intensity of characteristic x-rays for various elements as a function of polished thickness. Cu and Zn were detected only for the unpolished sample. It is reasonable that this Cu and Zn is deposited from the brass wire during EDM. For the samples with a polished thickness of more than ~10 μm the amount of Cu and Zn is almost zero as shown in figure 3 (b). This result is consistent with the EPMA measurement, i.e., the change in chemical composition is limited to ~ 10 μm from the sample surface after EDM.


141

Fig 3. (a) A photograph of one EDM cut SUS316L sample with different areas polished for 1, 2, and 5 min respectively. (b) The characteristic X-ray intensity of various elements normalized by the amount of Fe and plotted as a function of polishing thickness.

3.3 Results of XRD analysis

EDM cut samples were polished to various thicknesses and then analyzed by XRD. The residual stress determined from the XRD as a function of polishing thickness is shown in Figure 4. Here, the residual stress is the tensile stress [17]. Generally, the surface of a machined component has tensile residual stress after machining [18]. The residual stress decreased with increasing polished thickness from 0 to several tens μm and then became constant near zero level indicating that the defect layer induced by EDM is several tens μm thick.

Fig. 4. The polished depth dependence of the residual stress.


142

3.3 Results of PALS analysis PALS analysis were performed for samples with various polished thickness from 0 to ~ 100 μm. With increasing the

polished thickness, the positron lifetime becomes shorter which mean the defect induced by EDM decrease with increasing depth. The spectra and the fitting results of these spectra by using PALSFIT program will be presented on the conference.

4. Summary

We analyzed the defects near the surface of SUS316L induced by EDM by using electron microscope analysis, XRD, and PALS. The analyzed results are summarized below.

(1) Roughness of surface of the sample prepared by EDM is ~1.5 μm in standard deviation from the average level and is ~10 μm in the maximum amplitude from peak to valley.

(2) Chemical composition near surface (~ 10 μm) differs from that in bulk. In particular, elements from the EDM cutting wire (Cu and Zn) were detected within 10 μm depth.

(3) The residual tensile stress decreased with increasing polishing thickness suggesting defect layer with a thickness of up to several tens of micro meters.

(4) The positron lifetime decreased with increasing polishing thickness, indicating that the defect concentration decreases with increasing depth.

Acknowledgement

This work was supported by the Cabinet Office strategic innovation creative program innovative structural materials (SIP). We also acknowledge the support of Yoshihisa Harada and Kenji Ito of AIST for preparation of samples and RI-based positron lifetime measurement.

6. Reference

[1] A. K. Noor, Computer and structure 74, 507 (2000)

[2] J. C. Williams, Acta materialia 51, 5775 (2003)

[3] G Sposito, NDT & E International 43, 555 (2010)

[4] Y. Kawaguchi, Materials transaction 43, 727 (2002)

[5] J.H. Hartley, Applied surface Science 149, 204 (1999)

[6] F. Hori, Applied Surface Science 242, 304 (2005)

[7] P. Horodek , Tribology letters. 45, 341 (2012)

[8] H. Hansen, Physica satus solidi 69, 625 (1982)

[9] S. McNeil, Journal electronic materials 29, 583 (2003)

[10] F. Borner, Journal applied physics 84, 2255 (1998)

[11] A. Morris, Applied mechanics and materials 2006;5-6:145-52

[12] P. Hautojarvi. Positron in solid. Berlin: Springer Verlag; 1979.

[13] K.H. Ho, International Journal of Machine Tools & Manufacture 43,1287 (2003)

[14] U. Holzwarth, Applied physics A 73, 467 (2001)

[15] A. Yabuuchi, Journal of Nuclear Materials.419, 9 (2011)

[16] , 9 2011

[17] S. Kumar, Journal of Materials Processing Technology. 209, 3675 (2009)

[18] I.C. Noyan, J.B. Cohen, Residual Stress, Measurement by diffraction and interpretation, Springer-Verlag, New York,1987.


143

Life Prediction based on Damage Evaluation for Stainless Steel

Yoshihisa Harada1), Shuhei Kozu 2), Tokuo Teramoto 2), Nagayasu Oshima1) 1) National Institute of Advanced Industrial Science and Technology (AIST), 1-2-1 Namiki, Tsukuba, Ibaraki 305-8564,

TEL:029-861-7169/FAX:029-861-7853, [email protected] 2) Graduate School of Systems and Information Engineering, University of Tsukuba

Abstract: As fundamental research aim to fatigue life prediction for structural components, the effect of microstructure on mechanical properties in type 316L austenitic stainless steel deformed by static tensile or fatigue at room temperature. The tensile behaviors until 1.0 and 3.1% nominal strains were obtained associating with the deformation conditions. The fatigue test were conducted at ±0.2 % nominal strains and the strain rate of 0.1%/sec. Microstructural observations were carried out using electron back scatter diffraction pattern (EBSD). The misorientation parameters, such as the kernel averaged misorientation (KAM) were calculated by EBSD. Positron annihilation was measured as a non-destructive tool for the study of microstructural defects in this metal using the specimen deformed by 1.0 and 3.1% tensile strains.

1. Introduction In structural components for electric power plants, transportation and other specialized material function, the materials integrity and operational life prediction has been an important. Especially, residual stresses caused by plastic deformation in metal forming and welding processes may cause the stress corrosion cracking or deteriorate the fatigue properties. Also, thermal and mechanical stresses at operating temperatures may cause the life prediction by fatigue. Therefore, inspection method of plastic and fatigue deformation has been required as the standard damage and/or early fatigue damage estimation procedure for above structural components. Macroscopic fatigue properties are attributed to microstructures such as dislocation structures, grain boundaries, crystal orientations and precipitates [1,2].

In this study, as fundamental research aim to fatigue life prediction for structural components, the effect of microstructure on mechanical properties in type 316L stainless steel deformed by static tensile or fatigue at room temperature. The tensile behaviors until 1.0 and 3.1% nominal strains were obtained associating with the deformation conditions. The fatigue test were conducted at ±0.2 % nominal strains and the strain rate of 0.1%/sec. Microstructural observations were carried out using electron back scatter diffraction pattern (EBSD). The misorientation parameters, such as the kernel averaged misorientation (KAM) were calculated by EBSD. Positron annihilation was measured as a non-destructive tool for the study of microstructural defects in this metal using the specimen deformed by 1.0 and 3.1% tensile strains.

2. Experimental Procedure The material used in this study was solution heat-treated type 316L austenitic stainless steel. The chemical compositions and tensile properties of the steel are summarized in Tables 1 and 2. Round-bar tensile and fatigue specimens with the flat plate shape in the gage section (gage length=14 mm and cross section=6×10 mm2) were machined by electrical discharge machining (EDM) from the type 316L stainless steel plate. The static tensile tests carried out at a constant displacement rate of 1.0 mm/min until 1.0% or 3.1% nominal strain. For fatigue tests, the specimens were subjected to ±0.2 % nominal strains having a sinusoidal wave at the strain rate of 0.1%/sec and stress ratio of -1. The both tests were conducted using a closed-loop hydraulic MTS-810 testing machine of a dual servo valve type.

After tests, the specimen was machined by EDM in the shape of approximately 10×13×2 mm. The observed surface was polished using up to 0.1 μm alumina followed by colloidal silica polishing. After that, the electrolytic polishing was conducted using ethylene glycol-NaCl solution. Crystal orientations were measured using commercial software and EBSD detector interfaced to a field emission electron gun SEM (Carl Zeiss ULTRA PULS) operating at 15-20 keV. Tensile specimens were subjected to positron annihilation analysis at University of Tsukuba [3].

Table 1. Chemical composition of type 316L stainless steel (wt.%).

Fe C Si Mn P S Cr Ni Mo Bal. 0.0111 0.44 0.79 0.020 0.001 17.11 12.24 2.03

Table 2. Tensile properties of type 316L stainless steel.

Yield stress (MPa)

Tensile strength (MPa)

Elongation (%)

210 530 60


144

3. Results and Discussion 2.1 Relation between the EBSD parameter and plastic strain by tension Figure 1 shows stress-strain curve of type 316L stainless steel applied with the strain of 1.0% and 3.1%. These yield stress are around 210MPa as well as Table 2. After these specimen were machined by EDM and polished, crystal orientations were measured by EBSD. Figure 2 shows kernel average misorientation (KAM) [1] maps obtained from the EBSD measurements for the type 316L stainless steel tensile tested at the strain of (a) 0%, (b) 1.0% and (c) 3.1%. KAM is for a given data point the average misorientation between the data point and all of its neighbors is calculated exclude miorientations greater than some prescribed value 5º in this case. This local misorientation parameter can be calculated from the measured orientations and they correlate well with the degree of damage due to plastic strain, fatigue and creep [4,5]. Due to the plastic strain, the local misorientation develop inhomogeneously. Especially, the local misorientation tends to be large near grain boundaries. It seems that the magnitude of the local misorientation increases as the plastic strain increases.

Fig. 1 Stress-strain curve of type 316L stainless steel applied with strain of 1.0% and 3.1%.

(a) As-received (b) 1.0% strain (c) 3.1% strain

Fig. 2 Local misorientation maps obtained from the EBSD measurements for the type 316L stainless steel tensile tested at the strain of (a) 0%, (b) 1.0% and (c) 3.1%.

2.2 Relation between the EBSD parameter and fatigue damage Figure 3 shows typical stress-strain hysteresis curves of type 316L stainless steel under axial strain controlled low cycle fatigue. Since the maximum and the minimum strains were controlled, the peak stresses at positive and negative loads were changeable. In the case of Fig. 3, the maximum and the minimum strains of ± 0.2% were controlled and total strain range was kept to be 0.4%. It is clear that the peak stress increased at the initial stage and decreased after that gradually. The last cycle of 247114 reaches down to 25% load and stopped in the fatigue test. After this specimen was machined by EDM and polished, crystal orientations were measured by EBSD. Figure 4 shows KAM maps obtained from the EBSD measurements for the type 316L stainless steel in as-received (a) and (b) fatigue tested specimens. As

εε


145

well as Fig. 2, the local misorientation develop inhomogeneously. Figure 5 shows the local misorientation degree as a function of fraction in as-received and fatigue tested at ±0.2 % strain. The magnitude of the local misorientation increases as the fatigue damage increases.

Fig. 3 Stress-strain hysteresis curves of type 316L stainless steel under axial strain controlled low cycle fatigue.

(a) As-received (b) fatigue tested at ±0.2% strain Fig. 4 KAM maps obtained from the EBSD measurements for the type 316L stainless steel in as-received (a) and (b)

fatigue tested specimens.

Fig. 5 the local misorientation degree as a function of fraction in as-received and fatigue tested at ±0.2 % strain.


146

2.3 Relation between the S-parameter and plastic strain by tension Figure 6 shows the S-parameter of the as-received, 1% and 3.1% tensile strain specimens as a function of incident positron energy, E. It might be that the curve below 10 keV represents superposition of contributions from positron annihilating inside the sample and those which diffuse back and annihilate from a surface state due to machining and polishing. Above 10 keV, it can be predict that almost all positrons annihilate in the bulk of the sample. S-parameter gradually decreases with E above 10keV. As compared with as-received sample, tensile strain samples are large value of S-parameter due to increasing of defects such as dislocation density. On the other words, the positron annihilation is possible to measure degree of damage for fatigue and creep. It is under investigation.

Fig. 6 S-parameters as a function of incident positron energy E for as-received and tensile tested specimens.

3. Summary As fundamental research aim to fatigue life prediction for structural components, the effect of microstructure on mechanical properties in type 316L stainless steel deformed by static tensile or fatigue at room temperature. The tensile behaviors until 1.0 and 3.1% nominal strains were obtained associating with the deformation conditions. The fatigue test were conducted at ±0.2 % nominal strains and the strain rate of 0.1%/sec. Microstructural observations were carried out using electron back scatter diffraction pattern (EBSD). The misorientation parameters, such as the kernel averaged misorientation (KAM) tended to be large near grain boundaries. It seemed that the magnitude of the local misorientation increased as the plastic strain increased. It was indicated that the positron annihilation was possible to measure degree of damage for fatigue.

Acknowledgement The present work was partly supported by Cross-ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative measurement and analysis for structural materials (SIP-IMASM). We also acknowledge the support of Lixian Jiang, B.E. O’Rourke of AIST and Akira Uedono of university of Tsukuba for preparation of samples and RI-based positron lifetime measurement.

References [1] A.J. Schwartz, M. Kumar, B.L. Adams, D.P. Field, (Eds.) [Electron Backscatter Diffraction in Materials Science],

Springer, New York, (2009).

[2] J.H. Hartley, R.H. Howell, P. Asoka-Kumar, P.A. Sterne, D. Akers and A. Denison, “Positron annihilation studies of fatigue in 304 stainless steel”, Appl. Surf. Sci., 149, 204-206 (1999).

[3] A. Uedono, S. Ishibashi, N. Oshima, and, R. Suzuki, “Positron annihilation spectroscopy on nitride-based semiconductors”, Jpn. J. Appl. Phys. 52, 08JJ02 (2013).

[4] R. Yoda, T. Yokomaku and N. Tsuji, “Plastic deformation and creep damage evaluations of type 316 austenitic stainless steels by EBSD”, Mater. Charact , 61, 913-922 (2010).

[5] M. Muramatsu, T. Suzuki and Y. Nakasone, “Effets of microstructure on material properties of modified 9Cr-1Mo steel subject to creep-fatigue”, J. Mech. Sci. Tech., 29(1), 121-129 (2015).

ε

ε

ε


147

Two-component density functional study of positron-monovacancy interaction in metals

Shoji Ishibashi1) 1) Nanomaterials Research Institute, National Institute of Advanced Industrial Science and Technology (AIST),

1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan

Abstract: The positron-monovacancy interaction in d-block metals (except for Mn, Tc, and Hg), Mg and Al has been studied by the two-component density-functional-theory formalism. On the unrelaxed structure, the positron lifetime calculated with the presence of a positron is longer than that obtained neglecting the positron effect. In case that the positron effect is taken into account, the inward relaxation of the atoms surrounding the monovacancy is suppressed. Consequently, the difference is widened, especially for the group V metals. It has been found that the difference in the positron lifetime can be related to the bulk modulus and the cohesive energy.

1. Introduction The positron can be utilized as a powerful probe for detecting vacancy-type defects in various solids, since positrons are selectively trapped at (cation) vacancies in metals and semiconductors [1,2]. The momentum distribution of positron-annihilation radiation and the positron lifetime reflect the local environment of the annihilation site. This makes it possible to detect defects and to distinguish them. In practical cases, theoretical predictions are often crucial in interpreting experimental results and in identifying defect species. To describe the positron state in solids, where many nuclei and electrons exist, Boroński and Nieminen proposed the two-component density-functional-theory formalism [3]. In many practical calculations, a simplification is made assuming that the positron affects neither the electronic structure nor the atomic arrangement. This simplified scheme is called “conventional scheme”. To describe the delocalized positron state in a bulk, the conventional scheme is appropriate. Although the term “two-component scheme” (or two-component DFT) is sometimes used to describe the conventional scheme, here, it is used distinctively from the conventional scheme.

So far, there have been a limited number of applications of the (full) two-component scheme in calculating positron states trapped at defects. One reason is that the computational cost for the two-component scheme is approximately 10 times higher than that for the conventional scheme. Another reason is that, in many cases, the conventional scheme and the two-component scheme give similar annihilation parameters (Doppler broadening spectra and positron lifetimes) because of the feedback effect [4]. However, Saito and Oshiyama pointed out that there are significant differences between the positron lifetimes calculated by the two schemes the for vacancy clusters in Si [5]. For the two-component scheme, they used the version proposed by Gilgien et al. (the GGGC version) [6], where the electron-positron correlation energy is described in the dilute limit concerning the positron density. We calculated Doppler-broadening spectra and positron lifetimes using the version proposed by Puska, Seitsonen and Nieminen (the PSN version), where fully self-consistent calculations are performed using the density-dependent correlation energy, for divacancies in C, Si, Ge, SiC, AlN, GaN and InN and found that the difference between the two schemes depends on the bulk modulus [7]. For the relatively soft materials Si and Ge, the difference is significant.

In this study, a systematic computational study of the positron-monovacancy interaction in d-block metals (except for Mn, Tc, and Hg) [8], Mg and Al is presented. Positron states and lifetimes for a monovacancy in these metals have been calculated by the conventional scheme and the two-component scheme (the PSN version [6]) with and without structural relaxations.

2. Computational method All the calculations have been performed with our computational code QMAS (Quantum MAterials Simulator) [9-11]. In the conventional scheme, electronic structures, if necessary, and also atomic positions are calculated independently of the presence of the positron. On the other hand, in the two-component scheme, electronic structures and atomic positions (when they are relaxed) are affected by the presence of the positron. Electronic wave functions were described with the plane-wave basis in the framework of the projector augmented-wave (PAW) method [12]. Further details are described in Ref. [8]. 3. Results and discussion Positron lifetimes at a monovacancy have been calculated by the conventional (CV) or two-component (TC) scheme with no relaxation (NR) or relaxation (RL) and are plotted in Fig. 1 as a function of the atomic number Z. The overall tendency is similar to that for the positron lifetime in the bulk [13] as already pointed out by Campillo Robles et al. [14].


148

For d-block metals, the Z-dependence of the positron lifetime is thought to be attributed to the cohesion mechanism due to d electrons as mentioned in Ref. [8].

Fig. 1. Z-dependence of the positron lifetime τ trapped at a monovacancy. Results are calculated by the conventional (CV) or two-component (TC) scheme with no relaxation (NR) or relaxation (RL).

Fig. 2. Z-dependence of the difference in positron lifetimes trapped at a monovacancy Δτ calculated with no relaxation (NR) or relaxation (RL).

To clarify the positron effect, differences between the positron lifetimes obtained by the CV and TC schemes Δτ have been evaluated for both the NR and RL cases and are plotted in Fig. 2. Even for the NR case, where calculations by the CV and TC schemes were performed on the same structure for each element, there are significant differences between the two schemes. On each material except for Cu, it has been confirmed that the electron-positron overlap is larger for the TC scheme. The longer lifetime for the TC scheme is attributed to the lower enhancement factor than that of the dilute limit, which is used in the CV scheme. Larger differences are obtained for the RL case although the overall behavior is similar to that in the NR case except for the results of the group V BCC transition metals (V, Nb, and Ta). For these three elements, the differences in the RL case are remarkably large. It is also noted that the Z-dependence of Δτ resembles that of τ itself, implying the correlation with the cohesion property [9].


149

To explain the larger differences in the RL case, the change in the average distance Δd of the first-nearest-neighbor (FNN) (6 for HCP, 12 for FCC, 8 for BCC) atoms from the monovacancy center has been evaluated. For HCP metals, values for the six second-nearest-neighbor atoms have been obtained. If positrons do not exist, the FNN atoms show more or less inward relaxation. For the group V metals V, Nb, and Ta, the inward relaxation is obvious. In the presence of a positron, however, the inward relaxation is suppressed and sometimes outward relaxation occurs. The distinctive behavior of Δτ for V, Nb, and Ta in the RL case is attributed to the prominent difference in Δd between the CV and TC schemes. As already mentioned in our previous papers [7, 8], Δτ can be related to the bulk modulus and the cohesive energy.

It has been confirmed that the presence of a positron affects the monovacancy structure and the resultant positron lifetime. Calculations by the two-component DFT scheme are desirable for many cases, especially in dealing with the group V metals and soft (low bulk modulus) materials.

Acknowledgment

The author is grateful to Dr. Masanori Kohyama, Professor Akira Uedono, and Professor Kiyoyuki Terakura for useful information and stimulating discussions. This work was partly supported by the Cross-Ministerial Strategic Innovation Promotion Program - Unit D66 - Innovative Measurement and Analysis for Structural Materials (SIP-IMASM), operated by the Cabinet Office, Japan, and also by the Strategic Programs for Innovative Research (SPIRE), MEXT, and the Computational Materials Science Initiative (CMSI), Japan, under grant number hp140233.

References [1] M. J. Puska and R. M. Nieminen, “Theory of positrons in solids and on solid surfaces”, Rev. Mod. Phys. 66(3),

841-897 (1994).

[2] F. Tuomisto and I. Makkonen, “Defect identification in semiconductors with positron annihilation: Experiment and theory”, Rev. Mod. Phys. 85(4), 1583-1631 (2013).

[3] E. Boroński and R. M. Nieminen, “Electron-positron density-functional theory”, Phys. Rev. B 34(6), 3820-3831 (1986).

[4] M. J. Puska, A. P. Seitsonen and R. M. Nieminen, “Electron-positron Car-Parrinello methods: Self-consistent treatment of charge densities and ionic relaxations”, Phys. Rev. B 52(15), 10947-10961 (1995).

[5] M. Saito and A. Oshiyama, “Lifetimes of positrons trapped at Si vacancies”, Phys. Rev. B 53(12), 7810-7814 (1996).

[6] L. Gilgien, G. Galli, C. Gygi, and R. Car, “Ab Initio Study of Positron Trapping at a Vacancy in GaAs”, Phys. Rev. Lett. 72(20), 3214-3217 (1994).

[7] S. Ishibashi and A. Uedono, “Computational studies of positron states and annihilation parameters in semiconductors vacancy-type defects in group-III nitrides ”, J. Phys.: Conf. Ser., in press.

[8] S. Ishibashi, “Computational study of positron-monovacancy interaction in d-block metals”, J. Phys. Soc. Jpn. 84(8), 083703 1-4 (2015).

[9] http://qmas.jp.

[10] S. Ishibashi, T. Tamura, S. Tanaka, M. Kohyama and K. Terakura, “Ab initio calculations of electric-field-induced stress profiles for diamond/c-BN (110) superlattices”, Phys. Rev. B 76(15), 153310 1-4 (2007).

[11] S. Ishibashi and A. Uedono, “First-principles calculation of positron states and annihilation parameters for group-III nitrides”, J. Phys.: Conf. Ser. 505, 012010 1-4 (2014).

[12] P. Blöchl, “Projector augmented-wave method”, Phys. Rev. B 50(24), 17953-17979 (1994).

[13] I. K. MacKenzie, T. E. Jackman, and N. Thrane, “Positron Mean Lifetimes in Annealed Metals”, Phys. Rev. Lett. 34(9), 512-513 (1975).

[14] J. M. Campillo Robles, E. Ogando, and F. Plazaola, “Positron lifetime calculation for the elements of the periodic table”, J. Phys.: Condens. Matter 19(17), 176222 1-20 (2007).


150

Evaluation of Damage by Laser Cutting Process in CFRP

Shuhei Kozu1), Yoshihisa Harada 2), Hiroyuki Niino2), Tokuo Teramoto3) 1) Graduate School of Systems and Information Engineering, University of Tsukuba, 1-1-1 Tennodai, Tsukiuba-shi, Ibaraki 305-8573

[email protected] 2) National institute of Advanced Industrial Science and Technology (AIST)

3) Graduate School of Systems and Information Engineering, University of Tsukuba

Abstract: Cross-ply carbon fiber reinforced plastics (CFRP), which is expected to reduce the weight of transportations, was cut by a CW(continuous wave) fiber laser (λ=1.06μm, 2.0kW or 3.3kW) and machining with milling. Although the high-power laser cutting processes make it possible to improve cutting rate of CFRP, the laser-cut specimens clearly showed a thermal damage with a heat-affected zone (HAZ) identified by a micro X-ray CT analysis and infrared thermography.

1. Introduction

Carbon fiber reinforced plastic (CFRP) is the advanced materials attracting attention because of their specific strength and rigidity. In late years, for the environmental problems such as global warming, reducing the weight of transportation is important. CFRP is most attractive material to solve this problem, but does not reach the extensive use. One of the reasons includes difficulty workability. In general, processing method used for CFRP are mechanical-cutting and water-jet. These method has a problem about tool wear or water treatment, and the cutting speed using these method is limited [1]. Laser processing attracts attention as a method to overcome this problem because of possible high speed and no-tool wear. On the other hand, the laser processing remains thermal damage called Heat Affected Zone (HAZ) on CFRP. In the previous study, it has been revealed that HAZ influenced to mechanical properties of CFRP [2, 3]. However, there is no study about the detailed mechanism.

In this study, we gave laser processing for a cross-ply CFRP and investigated the influence of thermal damage and the degradation mechanism of mechanical properties by non-destructive inspection technology such as micro X-ray CT and infrared thermography.

2. Experimental procedure

The cross-ply CFRP [0/90]3S investigated in this study was consisted of 12 laminas called prepreg made by thermosetting epoxy resin and PAN-based carbon fibers. Laminates was composed of unidirectional fiber stacking 0 ° and 90° to the long direction distance of specimen. It was a thickness of approximately 3 mm. The specimen shape assumed it a rectangle of 25mm in width, 200mm in length with laser processing. The laser processing machine used in this study was continuous wave (CW) fiber laser (FL), wavelength 1.06μm, equipped with a high-speed scanner for cutting processing and max output power 6 kW. Table 1 shows processing condition. We changed the laser output, the cutting number of times and the cutting method on a condition. The Specimen 1 is the mechanical-cut specimen with endmills of cemented carbide to compare of the mechanical properties and the processing quality. After processing, the specimens were evaluated by micro X-ray CT and infrared thermography.


3.1 Thermal damage observation by micro X-ray CT Figure 1 (a), (b) and (c) show the cross-section images of specimens observed by micro X-rays CT. Fig.1 (a) shows that there is no change of the contrast on the cutting edge, so processing quality was high in the mechanical-cut specimen. Fig. 1 (b) and (c) show that there is a change of the contrast on the cutting edge. In addition, the thermal effect was controlled in 3.3 kW-FL processing specimen in comparison with the 2.0 kW-FL processing specimen. To cut high-quality and high-speed with laser processing, it is important matter for raising the output power of the laser processing machine and laser scanning speed.

Table 1 The condition of laser processing for cross-ply CFRP [0/90]3S

Specimen 1 (Mechanical-cut) 2 (2.0 kW-FL) 3 (3.3 kW-FL)

Average power [kW]

- 2.0 3.3

Cutting process - Single-line Multi-line

(Pitch : 1 mm) Cutting speed

[m/min] 0.12

3.0 (Scan speed : 0.2 m/s)

6.0 (Scan speed : 3.6 m/s)


151

Fig.1 Observation of the cross-section by micro X-ray CT

(a) mechanical-cut, (b) 2.0 kW-FL-cut and (c) 3.3 kW-FL-cut specimen

3.2 Thermal damage evaluation by infrared thermography Infrared thermography is non-destructive inspection technique that converts specimen surface temperature changed under cyclic stress to image mapping. Infrared thermography measurement can clarify with two method. One of the method is imaging the sum of principal stresses on the object surface distribution using thermoelastic effect. The other is damage evaluation from a minute temperature change to be caused by dissipated energy with the irreversible transformation of the material [4, 5].

For fatigue tests, the specimens were subjected to tensile cyclic loads having a sinusoidal wave at a frequency of 5 Hz and stress ratio of 0.1. Figure 2 shows the distribution image of the sum of principal stresses in mechanical-cut specimen and the 2.0 kW-FL processing specimen on the cutting edge near the center. Low stress distribution was seen in the region of approximately 1 mm in the 2.0 kW-FL processing specimen from the cutting edge. In this low stress distribution region, a result to be consistent with the size of the region of HAZ on micro X-rays CT. It is thought that the decrease of the load bearing region is the cause of the degradation of mechanical properties. Figure 3 shows the result that measured temperature changes due to the dissipated energy during fatigue test with the maximum cyclic stress of 720 MPa. In the 2.0 kW-FL processing specimen, temperature due to dissipated energy generated up and down frequently. It is thought that destruction of CFRP cut by the laser processing had been occurred on HAZ during cyclic loading, and affected the degradation of fatigue life.

Fig. 2 Thermo-elastic stress analysis for (a) mechanical-cut and (b) 2.0kW-FL-cut specimen

(Cyclic maximum load 8 kN)

1 mm 1 mm 1 mm

(a) Mechanical-cut (b) 2.0 kW-FL (c) 3.3 kW-FL

Measurement area

20 mm

16 mm

50

100

150

200

[MPa] 5 mm 5 mm

(a) Mechanical-cut (b) 2.0 kW-FL

Specimen


152

0

0.5

1

1.5

2

2.5

0.0001 0.001 0.01 0.1 1

Mechanical-cut2.0 kW-FL

Dis

sipa

ted

ener

gy [

K]

Fatigue life ratio , N/Nf

Fig. 3 Temperature change due to dissipated energy for the fatigue test in mechanical-cut and 2.0kW FL-cut specimens (Cyclic maximum stress 720 MPa)

4. Conclusion

In this study, the thermal damage of CFRP processed by advanced fiber laser cutting was evaluated by micro X-ray CT and infrared thermography. (1) By infrared thermography measurement, the low stress due to distribution image was formed on the cutting edge of laser processing specimen. This region was the cause with the degradation of mechanical properties. (2) From a temperature change due to dissipated energy during cyclic loading, destruction of CFRP cut by the laser processing had been occurred on HAZ during cyclic loading, and affected the degradation of fatigue life.

Acknowledgments This research was partially supported by High-Power Pulsed Fiber Laser and Processing Technology Project from New Energy and Industrial Technology Development Organization (NEDO) of Japan.

References

[1] Y. Maeda, [Advanced technology of carbon fiber], CMC Publishing, Japan, (2007).

[2] Y. Harada, K. Kawai, T. Suzuki and T. Teramoto, “Evaluation of Cutting Process on the Tensile and Fatigue Strength of CFRP Composites”, Mater. Sci. Forum, 706-709, 649-654 (2012).

[3] Y. Harada, M. Muramatsu, T. Suzuki, M. Nishino and H. Niino, “Influence of Laser Process on Mechanical Behavior during Cutting of Carbon Fiber Reinforced Plastic Composites”, Mater. Sci. Forum, 783-786, 1518-1523 (2014).

[4] T. Sakagami, “Thermoelastic analysis by infrared thermography”, J. Japan Weld. Soc., 72, 51-55 (2003).

[5] A. Akai, D. Shiozawa and T. Sakagami, “Dissipated Energy Evaluation for Austenitic Stainless Steel”, J. Soc. Mater. Sci., 62, 554-561 (2013)


153

Failure and Damage Evaluation of Laser-joint in CFRP-Stainless Steel

Gen NAKAMA1),2), Yoshihisa HARADA2), Hiroyuki NIINO2), Tokuo TERAMOTO3)

1) Graduate school of Systems and Information Engineering, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, 305-8573. 2) National Institute of Advanced Industrial Science and Technology (AIST), 1-2-1 Namiki, Tsukuba, Ibaraki, 305-8564.

TEL:029-861-7169/FAX:029-861-7853, [email protected] 3) Graduate School of Systems and Information Engineering, University of Tsukuba

Abstract: Carbon fiber reinforced thermoplastics (CFRTP), which is expected to reduce the weight of transportations and stainless steel (SUS304) were joined in the form of lap joint by a diode laser (3kW, CW(continuous wave), 0.8-1.2m/min) using insert materials of thermoplastic elastomer sheet. When the irradiation speed is 1.0m/min, the shear strength was the highest showing cohesive failure at the joining area. While, 0.8m/min and 1.2 m/min specimens were lower values than 1.0m/min specimens showing the interfacial failure. The fatigue strength was very similar results to the tensile shear strength. These laser-joint specimens were lower shear / fatigue strength than adhesive-joint specimens. From the results of thermo-elastic stress analysis using the thermography, the stress concentration was observed at the joining area just before the fatigue failure.

1. Introduction Recently, global warming and energy shortages have been concerned. Carbon fiber reinforced plastics (CFRP), which is superior strength and rigidity than metallic structural materials such as iron and aluminum, has attracted attention as a solution to the problem [1]. In general, a joining process of metal-plastic materials, CFRP and metal, has been usually performed using adhesive bonds or mechanical fasteners such as bolts and rivets. However, these joining processes lead to volatile organic compound (VOC) emission and inconvenience in mass-productions.

In this study, laser processing that can joint without contact is focused. The joining possibilities, strength and characteristics of the joint between carbon fiber reinforced thermoplastics (CFRTP) and SUS304 stainless steel using a cw diode laser were performed [2,3]. Comparing the shear strength from the tensile / fatigue test, and to visualize stress loading state of the material thermo-elastic stress analysis is applied to composite materials CFRTP.

2. Experimental details

2.1 Material The material used in this study is made of carbon fiber ( PAN-based chopped fiber about 0.2-7mm) and thermoplastic resin ( nylon6 ), it is called Carbon Fiber Reinforced Thermo-Plastic ; CFRTP. As stainless steel, SUS304 was used in this study. The joint specimens were prepared using flat stainless steel plates:100mm long, 50mm width and 2mm thick and CFRTP plates:100mm long, 50mm width and 3mm thick. In this case, two kinds of method for joining were prepared, adhesive joint and laser joint by direct diode laser (3kW, CW, processing speed:0.8-1.2m/min).

2.2 Tensile, fatigue test The tensile and fatigue tests were performed on the servo hydraulic testing machine MTS810 from MTS systems Gmbh. In the tensile test, the specimens were pulled up to the fracture at a rate of 0.5mm/min. In the fatigue test, the specimens were applied cyclic loads at a frequency of f=5Hz with a maximum number of cycles of N=107.

2.3 Therm-oelastic stress analysis method Fatigue test was subjected with a load in the elastic range from the graph of tensile shear strength. During fatigue test, it was subjected to thermal-elastic stress analysis using infrared thermography [4]. For this analysis, an infrared thermography ( FLIR; SC7500 ) and lock-in analysis software ( FLIR; Altair LI ) was used in this study. Frequency of the camera is 200Hz, and 4000 images were gotten by photographing 20 seconds.


3.1 Shear strength and fatigue strength Figure 1 shows the shear strength obtained by the tensile test of the laser-joint specimen with the irradiation speed of 0.8, 0.9, 1.0, 1.1, 1.2 m/min. It also shows the results of the adhesive-joint material in the figure. The laser-joint specimens is lower strength to adhesive joint specimens. While, it can be seem that the difference of shear strength due to the irradiation speed. Therefore, there is optimum parameter for laser-joint process. When the irradiation speed is 1.0m/min, the shear strength was the highest showing cohesive failure at the joining area. While, 0.8m/min and 1.2 m/min specimens were lower values than 1.0m/min specimens showing the interfacial failure. It is considered that the optimal irradiation conditions in this laser is 1.0m/min.


154

Figure 2 shows S-N curves for fatigue test in laser-joint and adhesive-joint specimens. Fatigue strength is seen a similar trend to figure of shear strength. In the irradiation rate of 1.0m/min, the highest strength was shown. It can be said that because of the strength is lowered by other radiation rate, the optimum value of the irradiation rate is 1.0m/min.

Fig.1 Shear strength of laser-joint and adhesive-joint specimens

Fig.2 Fatigue strength of laser-joint and adhesive-joint specimens

3.2 Damage analysis Figure 3 shows the thermos-elastic stress analysis image of laser-joint specimen obtained by the infrared thermography. The image subjected to fatigue test for the specimens that were joined by laser irradiation speed 1.0m/min, and those obtained at the time of 1,000cycles and 20,000cycles. The test condition is the cyclic maximum stress, σ=6[MPa], stress ratio, R=0.1, and the frequency, f=5[Hz]. Focusing on the joint area, fatigue initially is almost evenly stress is loaded, but before fracture can be seen that extremely high stress is loaded in the near the center and right side of the fracture surface. From this fact, this specimen is considered to have destroyed from the stress concentration section as shown in Fig.3.


(a) Initial fatigue 1000cycles

(b) Just before faliure 20000cycles

Fig.3 Thermo-elastic stress analysis of laser-joint specimen (1.0m/min)

4. Conclusions

Carbon fiber reinforced thermoplastics (CFRTP) and stainless steel (SUS304) were joined in the form of lap joint by a semiconductor laser (3kW, CW, 0.8-1.2m/min) using insert materials of thermoplastic elastomer sheet. When the irradiation speed is 1.0m/min, the shear strength achieved 10MPa and revealed an appropriate irradiation conditions. While, 0.8m/min and 1.2 m/min specimens were lower values than 1.0m/min specimens showing the interfacial failure. The fatigue strength was very similar results to the tensile shear strength. These laser-joint specimens were lower shear / fatigue strength than adhesive-joint specimens. From the results of thermo-elastic stress analysis using the thermography, the stress concentration was observed at the joining area just before the fatigue failure.

Acknowledgments This research was supported by High-Power Pulsed Fiber Laser and Processing Technology Project from New Energy and Industrial Technology Development Organization (NEDO) of Japan.

References [1] Y. Maeda, “The Recent Trends of Carbon Fiber ( in Japanese)”, CMC Publishing, Japan, (2007).

[2] Y. Harada, K. Kawai, T. Suzuki, T. Teramoto, “Evaluation of Cutting Process on the Tensile and Fatigue Strength of CFRP Composites”, Mater. Sci. Forum, 706-709, 649-654(2012).

[3] Y. Harada, K. Kawai, M. Nishino, H. Niino, T. Suzuki, T. Teramoto, “Effect of Fiber Orientation on Tensile Properties of Carbon Fiber Reinforced Plastic (CFRP) using Laser Cutting Process”, Proc. 12th Jpn Inter. SAMPE Sympo. Exhibit., (2011).

[4] J. Motesano, Z. Fawaz, H. Bougherara, “Use of infrared thermography to investigate the fatigue behavior of a carbon fiber reinforced polymer composite”, Comp. Struct., 97, 76-83(2013).


156

Bus stops

Free shuttle bus from AIST to Tsukuba and Arakawaoki stations


157


158

City bus from Namiki-2 chome to Tsukuba and Arakawaoki stations


159

Timetable of the Tsukuba Express (TX) and JR lines


160

Lunch guide Please ask the Japanese staff for details.


161

Annual report 2015 of Innovative Measurement and Analysis for Structural Materials 1st Symposium on Innovative Measurement and Analysis for Structural Materials

Published on September 29, 2015 Editor

National Institute of Advanced Industrial Science and Technology (AIST) 1-1-1, Umezono, Tsukuba, Ibaraki, 305-8568, Japan Committee for the 1st Symposium on Innovative Measurement and Analysis for Structural Materials TEL: 029-861-5685, URL: https://staff.aist.go.jp/m.ohkubo/SIP-IMASM e-mail: [email protected] M. Ohkubo, Y. Harada, P. Fons, T. Sakaguchi, A. Miyoshi, and N. Koizumi

No part of this report should be reproduced or copied without permission from the editor and the authors.

2015

1

2015 9 29

305-8568 1-1-1

1

TEL: 029-861-5685, URL: https://staff.aist.go.jp/m.ohkubo/SIP-IMASM e-mail: [email protected]


KEK


163

AIST15-C00008

Documents

Annual report 2015 of Innovative Measurement and …...The First Symposium on SIP Innovative measurement and analysis for structural materials (SIP-IMASM 2015) 5 3-2 Kazuhiro Hono